analysis of nike distribution facility’s outbound process flow

Analysis of Nike Distribution Facility’s Outbound

Submitted in partial fulfilment of the requirements

BACHELORS OF INDUSTRIAL

FACULTY OF ENGINEERING, BUILT ENVIRO

UNIVERSITY OF PRETORIA


Process Flow

by

Werner Wagenaar

26014964

Submitted in partial fulfilment of the requirements

for the degree of

BACHELORS OF INDUSTRIAL ENGINEERING

in the

FACULTY OF ENGINEERING, BUILT ENVIRONMENT

AND INFORMATION

UNIVERSITY OF PRETORIA

PRETORIA

October 2009


MENT

Executive Summary

Nike is proud to be associated with the Soccer World Cup 2010 event through the sponsorship

of footwear and kits to various teams and high profile players. Being a leading global brand

Nike is expecting extraordinary demand over the duration of the event. The agility and flexibility

in which Nike respond to these special orders is what will distinguish them f

in this space. The service levels expected by customers for World Cup related orders add

significant pressure to the entire supply chain. The Nike South Africa distribution facility, located

in Meadowdale, is currently not operating at it

flexibility to meet the customer requirements. Barloworld Logistics is contracted to analyse all

distribution processes and develop the capabilities needed to achieve and maintain order

fulfilment throughout the event.

A simulation study was conducted

affected by this event. Along with

vital in the decision making of the build up for the World

not all Key Performance Indicators (

planning had to be reviewed to achieve the desired outcomes

was resistant to change and modification

flexibility.

The change in the resources deployment will save

productivity and utilisation in various

cycle time and order lead time, which in turn will result in increased customer satisfaction.

Increasing the flexibility indicated that more goods could be delivered in the same period of

time, thus increasing overall order fulfilment

Overall the project was a huge succe

yield significant cost reductions.

on unit processing costs and increase throughput capacity.

internally and further studies can be conducted on outbound process flow and planning

processes.


kits to various teams and high profile players. Being a leading global brand


in which Nike respond to these special orders is what will distinguish them from the competitors



in Meadowdale, is currently not operating at its optimal capacity and desired processing



A simulation study was conducted using Arena® to run different scenarios on process

affected by this event. Along with Microsoft Excel® spreadsheets this simulation proved to be

vital in the decision making of the build up for the World Cup. The results obtained showed that

Key Performance Indicators (KPI’s) where at their optimal levels and resource capacity

planning had to be reviewed to achieve the desired outcomes. It also indicated that the system

modifications on process logic are needed to create the desired

The change in the resources deployment will save Nike labour cost due to

various departments. The system changes will reduce process

rder lead time, which in turn will result in increased customer satisfaction.


overall order fulfilment.

s a huge success and once implemented in February 2010, the project will

Preliminary indications are that the project will yield a reduction

on unit processing costs and increase throughput capacity. The simulation model is re

nternally and further studies can be conducted on outbound process flow and planning


kits to various teams and high profile players. Being a leading global brand,


m the competitors



s optimal capacity and desired processing



to run different scenarios on processes

spreadsheets this simulation proved to be

ults obtained showed that

s and resource capacity

. It also indicated that the system

d to create the desired

cost due to increased

. The system changes will reduce process

rder lead time, which in turn will result in increased customer satisfaction.


ry 2010, the project will

Preliminary indications are that the project will yield a reduction

The simulation model is re-usable

nternally and further studies can be conducted on outbound process flow and planning

Acknowledgments

TO MY FAMILY AND FRIENDS

ANDRÉ HOUGH

PROF. PAUL KRUGER – PROJECT MENTOR AT

TO MY FAMILY AND FRIENDS – FOR THEIR UNCONDITIONAL LOVE AND SUPPORT

É HOUGH – BARLOWORLD LOGISTICS

PROJECT MENTOR AT THE UNIVERSITY OF PRETORIA

FOR THEIR UNCONDITIONAL LOVE AND SUPPORT

THE UNIVERSITY OF PRETORIA

Table of Contents

1 Introduction ................................

1.1 Background................................

1.1.1 Overview of Outbound Process Flow

1.2 Project Aim ................................

1.3 Project Scope ................................

2 Literature Study................................

2.1 System Analysis Models

2.1.1 Management Decision

2.1.2 Nature of Models ................................

2.2 Basic Approach to Modelling

2.3 Types of Models ................................

2.3.1 Physical Models ................................

2.3.2 Computer Simulation Models

2.3.3 Analytical Models ................................

2.3.4 Analytical versus Simulation Models

2.4 Why Model? ................................

2.5 Simulation Modelling ................................

2.5.1 Simulation Language

2.5.2 High-Level Simulators

2.5.3 Simulation Modelling Tools

3 Conceptual Design of Model

3.1 Structuring the Model ................................

3.2 Conceptual Inputs and Outputs of the Model

i

.........................................................................................................................

................................................................................................

Overview of Outbound Process Flow ................................................................

................................................................................................

................................................................................................

................................................................................................

System Analysis Models ..............................................................................................

Management Decision ..........................................................................................

................................................................................................

Basic Approach to Modelling ........................................................................................

................................................................................................

................................................................................................

Computer Simulation Models ................................................................

................................................................................................

Analytical versus Simulation Models ................................................................

................................................................................................

................................................................................................

Simulation Language ...........................................................................................

Level Simulators ..........................................................................................

Simulation Modelling Tools ................................................................

Conceptual Design of Model ..............................................................................................

................................................................................................

Conceptual Inputs and Outputs of the Model ..............................................................

......................... 1

.................................................. 1

.................................... 3

.................................................. 4

.............................................. 7

................................................... 8

.............................. 8

.......................... 8

.................................. 8

........................ 9

.........................................12

...................................12

...............................................12

.................................13

....................................13

................................................14

...................................15

...........................17

..........................17

..................................................17

..............................22

..................................22

..............................24

4 Data Collection and Analysis

4.1 Input Data and Analysis

4.1.1 Order Information ................................

4.1.2 Transport Time of a Picker per Order

4.1.3 Quantity of a Business Unit Picked

4.1.4 Number to be batched

4.1.5 Distance Conveyor travel in Picking

4.1.6 Process Time Picking

4.1.7 Process Time Packing

4.2 Soft Data ................................

5 Depiction of Simulation Model

5.1 Conceptual Translation

5.1.1 Entities ................................

5.1.2 Attributes ................................

5.1.3 Variables ................................

5.1.4 Resources ................................

5.1.5 Queues ................................

5.1.6 Conveyors ................................

5.1.7 Picker Travel Pattern

6 Simulation Model Construction

6.1 Model Activity Areas ................................

6.1.1 Order Information Generation

6.1.2 Pick Face ................................

6.1.3 Pack Station ................................

6.1.4 Despatch and Output transfer

ii

Data Collection and Analysis..............................................................................................

Input Data and Analysis ..............................................................................................

................................................................................................

Transport Time of a Picker per Order ................................................................

Quantity of a Business Unit Picked ................................................................

Number to be batched .........................................................................................

Distance Conveyor travel in Picking ................................................................

Process Time Picking ..........................................................................................

Process Time Packing .........................................................................................

................................................................................................

Depiction of Simulation Model ............................................................................................

Conceptual Translation ...............................................................................................

................................................................................................

................................................................................................

................................................................................................

................................................................................................

................................................................................................

................................................................................................

Picker Travel Pattern ...........................................................................................

Simulation Model Construction ...........................................................................................

................................................................................................

Order Information Generation ................................................................

................................................................................................

................................................................................................

and Output transfer ................................................................

..............................27

..............................27

.................................27

...................................28

......................................30

.........................30

.....................................31

..........................34

.........................35

.....................................................36

............................37

...............................37

.................................................37

.............................................37

..............................................39

...........................................40

................................................41

...........................................42

...........................42

...........................44

...................................44

..............................................44

.............................................47

........................................50

..............................................52

6.2 Verification and Validation of the Model

6.2.1 Verification ................................

6.2.2 Validation ................................

6.3 Scenario Development ................................

6.3.1 Increased growth in forthcoming years

6.3.2 Change in the number of resources

6.3.3 Change the number of orders released from the pick pool

6.4 Possible follow-up Scenarios

6.4.1 Business unit locations

6.4.2 Conveyor Speed ................................

7 Simulation Results and Recommendations

7.1 Increase growth in forthcoming years

7.1.1 Scenario 1: A ................................

7.1.2 Scenario 1: B ................................

7.1.3 Scenario 1: C ................................

7.2 Change in resource capacity

7.2.1 Scenario 2: A ................................

7.2.2 Scenario 2: B ................................

7.2.3 Scenario 2: C ................................

7.3 Change the batch size of orders send from the pick pool

8 Conclusion ................................

9 References ................................

10 Appendices ................................

10.1 Input Data Analysis ................................

10.2 Probability Distributions

iii

Verification and Validation of the Model ................................................................

................................................................................................

................................................................................................

................................................................................................

Increased growth in forthcoming years................................................................

Change in the number of resources ................................................................

Change the number of orders released from the pick pool ................................

up Scenarios ................................................................

Business unit locations ........................................................................................

................................................................................................

Simulation Results and Recommendations ................................................................

Increase growth in forthcoming years................................................................

................................................................................................

................................................................................................

................................................................................................

n resource capacity .......................................................................................

................................................................................................

................................................................................................

................................................................................................

he batch size of orders send from the pick pool ................................

.........................................................................................................................

........................................................................................................................

................................................................................................

................................................................................................

Probability Distributions ..............................................................................................

......................................53

...........................................53

.............................................54

................................55

.................................55

.....................................56

...................................56

......................................................57

........................57

..................................57

........................................58

..........................................58

.......................................58

.......................................59

.......................................61

.......................63

.......................................63

.......................................64

.......................................66

............................................68

.........................70

........................71

....................................................73

.....................................73

..............................77

10.2.1 Picking Process Time for Business Unit “Other” Distribution

10.2.2 Picking Process Time for Business Unit “Footwear” Distribution

10.2.3 Packing Process Time Distribution

10.3 Simulation Results ................................

10.3.1 Queue average waiting time per entity

10.3.2 Number of entities entering a process

10.4 The Simulation Parts ................................

10.4.1 Order information generation phase

10.4.2 Pick Face ................................

10.4.3 Pack Station ................................

10.4.4 Despatch and Output transfer phase

iv

Picking Process Time for Business Unit “Other” Distribution ................................

Picking Process Time for Business Unit “Footwear” Distribution ..........................

Packing Process Time Distribution ................................................................

................................................................................................

Queue average waiting time per entity ................................................................

Number of entities entering a process ................................................................

................................................................................................

Order information generation phase ................................................................

................................................................................................

................................................................................................

tch and Output transfer phase ................................................................

................................77

..........................78

.......................................79

......................................80

.................................80

..................................81

...................................82

.....................................82

.............................................83

........................................84

...................................85

List of Figures

Figure 1: Process Flow Diagram

Figure 2: Outbound Process Flow

Figure 3: Scope Boundary ................................

Figure 4: Basic Approach to Modelling

Figure 5: Areas influenced by modelling

Figure 6: Illustrating the mapping of simulation between the real world and models

Figure 7: Major Phases in Simulation S

Figure 8: Flow chart of an Arrival

Figure 9: Walking time from one block to next

Figure 10: Illustration of the split of an Order

Figure 11: Illustration of Conveyor Movement on Picking Levels

Figure 12: Level Layout ................................

Figure 13: Data input to Simulation Study

Figure 14: Picker Travel Pattern ................................

Figure 15: Read Simulation Model Variables

Figure 16: Duplication of Entity to the Number of Orders

Figure 17: Assigning and Reading Attributes to Entity

Figure 18: Assigning Orders to different l

Figure 19: Picking and Conveying Processes

Figure 20: Decide on packing station

Figure 21: Packing Process ................................

Figure 22: Delay, Batching and Output transfer processes

Figure 23: Despatching of orders

Figure 24: Resource utilisation of base model with 8 pickers

Figure 25: Resource utilisation of three year projection

Figure 26: Resource utilisation of six year projection

Figure 27: Resource utilisation with increase of three pickers

Figure 28: Resource utilisation with increase of seven pickers

Figure 29: Resource utilisation with decrease of three packing stations

v

Figure 1: Process Flow Diagram ................................................................................................

Figure 2: Outbound Process Flow ..............................................................................................

................................................................................................

Figure 4: Basic Approach to Modelling ................................................................

Figure 5: Areas influenced by modelling................................................................


Figure 7: Major Phases in Simulation Study ................................................................

Figure 8: Flow chart of an Arrival...............................................................................................

Figure 9: Walking time from one block to next ................................................................

Figure 10: Illustration of the split of an Order ................................................................

Figure 11: Illustration of Conveyor Movement on Picking Levels ................................

................................................................................................

Figure 13: Data input to Simulation Study ................................................................

................................................................................................

Figure 15: Read Simulation Model Variables ................................................................

Figure 16: Duplication of Entity to the Number of Orders ..........................................................

Figure 17: Assigning and Reading Attributes to Entity ...............................................................

Figure 18: Assigning Orders to different levels for picking .........................................................

Figure 19: Picking and Conveying Processes ................................................................

Figure 20: Decide on packing station ........................................................................................

................................................................................................

Figure 22: Delay, Batching and Output transfer processes .......................................................

Figure 23: Despatching of orders ..............................................................................................

Resource utilisation of base model with 8 pickers ................................

Figure 25: Resource utilisation of three year projection .............................................................

Figure 26: Resource utilisation of six year projection ................................................................

Figure 27: Resource utilisation with increase of three pickers ................................

esource utilisation with increase of seven pickers ................................

Figure 29: Resource utilisation with decrease of three packing stations ................................

................................ 3

.............................. 4

......................................... 7

......................................................11

....................................................15

Figure 6: Illustrating the mapping of simulation between the real world and models ..................16

..............................................23

...............................25

...........................................29

.............................................31

...............................................32

.............................................33

.................................................36

................................43

............................................45

..........................46

...............................47

.........................48

...........................................49

........................50

......................................51

.......................52

..............................53

....................................................59

.............................60

................................62

...................................................64

..................................................65

....................................67

Figure 30: Decrease in orders send from pick pool results

Figure 31: Resource utilisation with decrease in orders send from pick pool

List of Tables

Table 1: Distance Conveyor travel from each block

Table 2: Processing Times for Picking in sec/unit

Table 3: Processing Times for Packing

Table 4: List of Attributes ................................

Table 5: List of Variables ................................

Table 6: List of Resources ................................

Table 7: List of Queues ................................

Table 8: Conveyors ................................

Table 9: Base model results ................................

Table 10: Three year projection in quantity of units ordered results

Table 11: Six year projection in quantity of units ordered results

Table 12: Increase picker resource capacity by three results

Table 13: Increase picker resource capacity by seven results

Table 14: Decrease packer resource capacity by three results

Table 15: Picker Transport Time Table fo

Table 16: Picker Transport Time Table for Level B, C, D

Table 17: Extraction from Total Transport Time for Each Level in Picking

Table 18: Extraction from Quantity and Business Units per Order on each Level

Table 19: Extraction from Number to be batched on a Level

Table 20: Extraction Distance the Conveyor travel on a Level

vi

Figure 30: Decrease in orders send from pick pool results ........................................................

Figure 31: Resource utilisation with decrease in orders send from pick pool .............................

Table 1: Distance Conveyor travel from each block ................................................................

Picking in sec/unit ................................................................

Table 3: Processing Times for Packing ................................................................

................................................................................................

................................................................................................

................................................................................................

................................................................................................

................................................................................................

................................................................................................

n in quantity of units ordered results................................

Table 11: Six year projection in quantity of units ordered results ................................

Table 12: Increase picker resource capacity by three results ................................

Table 13: Increase picker resource capacity by seven results ................................

Table 14: Decrease packer resource capacity by three results ................................

Table 15: Picker Transport Time Table for Level A ................................................................

Table 16: Picker Transport Time Table for Level B, C, D ..........................................................

Table 17: Extraction from Total Transport Time for Each Level in Picking ................................

Table 18: Extraction from Quantity and Business Units per Order on each Level

Table 19: Extraction from Number to be batched on a Level ................................

Table 20: Extraction Distance the Conveyor travel on a Level ................................

........................68

.............................69

..................................34

.....................................35

.....................................................35

..........................................38

...........................................39

........................................40

.............................................41

...................................................42

......................................58

...........................................60

...............................................61

....................................................63

...................................................65

.................................................66

...................................73

..........................74

.................................75

Table 18: Extraction from Quantity and Business Units per Order on each Level ......................75

.....................................................76

...................................................76

1 Introduction

1.1 Background

Nike, Inc. is a global company operating in 4 regions and 6 continents and with global revenue

exceeding US $14 billion in 2006. Nike’s global headquarters are located in Beaverton in the

USA. South Africa is part of the EMEA (Europe, Middle

headquartered in Hilversum, the Netherlands.

Originally managed by a distributor, Nike decided to manage marketing and distribution in South

Africa through its own organisation which was implemented in 1995. The office is loc

Midrand, Johannesburg with a satellite sales office in Cape Town. The total organisation

employs 120 people with a turnover of approximately

classified in 3 categories:

� Footwear

� Categories such as Running, Basketb

� Apparel

� Products like Shirts, Tank Tops, Jerseys, Track Suits, Windbreakers, Sports

Jackets and Caps

� Equipment

� Items such as Socks, Bags, Back Packs, Gloves, Timing and Balls

Nike enjoys major market shares

Adidas, Ellesse and Puma in this space.

1



USA. South Africa is part of the EMEA (Europe, Middle-East and Africa) region which is

headquartered in Hilversum, the Netherlands.


Africa through its own organisation which was implemented in 1995. The office is loc

, Johannesburg with a satellite sales office in Cape Town. The total organisation

with a turnover of approximately $100m. Nike’s product

Categories such as Running, Basketball, Football, Cross-Training and Tennis

Products like Shirts, Tank Tops, Jerseys, Track Suits, Windbreakers, Sports

Jackets and Caps

Items such as Socks, Bags, Back Packs, Gloves, Timing and Balls

shares across all 3 business units’. They compete

in this space.



t and Africa) region which is


Africa through its own organisation which was implemented in 1995. The office is located in

, Johannesburg with a satellite sales office in Cape Town. The total organisation

product range can be

Training and Tennis

Products like Shirts, Tank Tops, Jerseys, Track Suits, Windbreakers, Sports

Items such as Socks, Bags, Back Packs, Gloves, Timing and Balls

They compete with Reebok,

In South Africa, product reaches customers through the following distribution channels:

• National and strategic accounts of multi

• Franchised Nike only stores

• Nike owned Factory outlets selling close

inventory

Nike’s distribution facility is loca

Logistics was contracted by Nike to investigate the local di

expected increase of sales during the 2010 Soccer World Cup in South Africa. I

that at least 50,000 units per day must be distributed through the distributions facility, to meet

the expected demand. The distr

30,000 units per day, which is far less than the projected need. Nike is a major sponsor of the

Soccer World Cup, which places an important responsibility on the distribution facility to run

its optimal level, thereby insuring a constant av

reduced share in the market to its various competitors. Nike’s ability to respond to a rapid

increase and decrease in demand is an important factor in its

competitors and retain market share.

The focus of Barloworld Logistics is on the distribution facility and all the processes it entails.

Thus an overview is needed of the facility and the recommendations to ensure that the

will be achieved for the Soccer World Cup 2010

2


National and strategic accounts of multi-brand stores

stores

Nike owned Factory outlets selling close-out (also called aged) inventory and excess

Nike’s distribution facility is located in Meadowdale, Johannesburg. During 2009

by Nike to investigate the local distribution facility in view of an

expected increase of sales during the 2010 Soccer World Cup in South Africa. I

000 units per day must be distributed through the distributions facility, to meet

the expected demand. The distribution facility is currently running at a level of approximately

000 units per day, which is far less than the projected need. Nike is a major sponsor of the

Soccer World Cup, which places an important responsibility on the distribution facility to run

its optimal level, thereby insuring a constant availability of products, failing


increase and decrease in demand is an important factor in its ability to compete with the main

competitors and retain market share.


Thus an overview is needed of the facility and the recommendations to ensure that the

will be achieved for the Soccer World Cup 2010 season.


inventory and excess

During 2009 Barloworld

stribution facility in view of an

expected increase of sales during the 2010 Soccer World Cup in South Africa. It is estimated

000 units per day must be distributed through the distributions facility, to meet

at a level of approximately

000 units per day, which is far less than the projected need. Nike is a major sponsor of the

Soccer World Cup, which places an important responsibility on the distribution facility to run at

might lead to a


ability to compete with the main


Thus an overview is needed of the facility and the recommendations to ensure that the demand

1.1.1 Overview of Outbound

In the distribution facility there are mainly two types of process flows. The first is the Inbound

Process Flow (Figure 1); this process flow takes into account all the process

and resources in receiving of goods from the suppliers. The departments running the Inbound

Process Flow are the Goods Rec

and the High Bay Department (the storing of the goods in bulk

The second process flow is the Outbound Process Flow (

account all the processes, flow of material and resources in despatching specific orders to

customers. The departments running the Outbound Process Flow are

Value Added Service (VAS), Packing department and the Despatching department.

Figure 1: Process Flow Diagram

On a higher level of detail the Outbound Process Flow consists of an order that comes from a

customer. This order contains all the necessary detail from the products to be purchased to the

delivery date of the products. After the order is complete it goes to the Pick Pool

From the Pick Pool it is taken to the Wave planning, where the planning occurs of when each

order must be picked in order to reach the specified delivery date.

constraints into account from the

delivery plan) which is the less urgent orders.

3

Overview of Outbound Process Flow

there are mainly two types of process flows. The first is the Inbound

; this process flow takes into account all the processes


he Goods Receiving department (the receiving of goods from the supplier)

and the High Bay Department (the storing of the goods in bulk storage locations

The second process flow is the Outbound Process Flow (Figure 1). This process flow takes into


customers. The departments running the Outbound Process Flow are the Fine P




of the products. After the order is complete it goes to the Pick Pool


picked in order to reach the specified delivery date. This planning takes di

from the, expedite, which is the most urgent to the Non IDP (Integrated

urgent orders. These orders get released by Wave planning and

there are mainly two types of process flows. The first is the Inbound

es, flow of material


of goods from the supplier)

storage locations).

his process flow takes into


Pick department,




of the products. After the order is complete it goes to the Pick Pool (Figure 2).


planning takes different

expedite, which is the most urgent to the Non IDP (Integrated

These orders get released by Wave planning and

Orders

Pick Pool

Wave

Planning

the picking of the goods takes place. After the goods have been picked it is sent to th

department. At the Packing Department the goods are packed into boxes and sent to the

Despatch department where it gets despatched to the

Figure 2: Outbound Process Flow

1.2 Project Aim

The main objective of this project is to provide Barloworld Logistics with a model to assess if the

outbound process flow of the distribution facility would meet the output and flexibility

requirements over the Soccer W

achieved for the increased sales o

any, which might take place at the distribution facility to run at optimum throughput capacity.

The facility was designed to achieve a theoretical output of 50,000

due to the expected high demand

be as close as possible to the theoretical output per day. The process analysis will include

demand variability, seasonality and

important and peak demand period.

4

Release

Picking

Packing

Despaching

the goods takes place. After the goods have been picked it is sent to th

At the Packing Department the goods are packed into boxes and sent to the

Despatch department where it gets despatched to the specific customer (Figure

: Outbound Process Flow

of this project is to provide Barloworld Logistics with a model to assess if the

of the distribution facility would meet the output and flexibility

World Cup period. It will also assess if the demand levels can be

achieved for the increased sales over the Soccer World Cup season, as well as any changes, if


as designed to achieve a theoretical output of 50,000 units per day. It is essential

due to the expected high demand of the Soccer World Cup that the actual output per day must


demand variability, seasonality and growth over the next few months, which

demand period.

Despaching

the goods takes place. After the goods have been picked it is sent to the Packing

At the Packing Department the goods are packed into boxes and sent to the

Figure 2).

of this project is to provide Barloworld Logistics with a model to assess if the

of the distribution facility would meet the output and flexibility

. It will also assess if the demand levels can be

, as well as any changes, if


units per day. It is essential

of the Soccer World Cup that the actual output per day must


growth over the next few months, which are Nike’s most

An effective response to the changes in customer demand is the decisive factor. The demand

will increase during the build-up and the occurrence of the Soccer World Cup 2010. With

continuous changes comes a need for intuitive and fle

great importance for the Soccer World Cup. If a team goes through to the n

Soccer World Cup, it is expected that the demand for

They will be ordered today and need to be

orders are of higher priority and the flexibility of

met.

The simulation model will be used in studying accurate information and recommendati

management pertaining to the following decision making parameters:

• Highlight the possible bottlenecks under a certain set of conditions

• Flexibility of the system under changes in demand

• Material flow in the distribution facility

• Number of staff needed at each station

• Increase throughput, as close as possible to the theoretical maximum per day

• Calculate the buffer sizes, if needed, at each station

• Increase resource utilisation

• Inter relationship between processes

5


up and the occurrence of the Soccer World Cup 2010. With

continuous changes comes a need for intuitive and flexible control. A highly flexible

great importance for the Soccer World Cup. If a team goes through to the n

it is expected that the demand for Nike products of that team will increase.

ay and need to be despatched the next day, this indicates that some

and the flexibility of the system will indicate if the demand can be

The simulation model will be used in studying accurate information and recommendati

management pertaining to the following decision making parameters:

Highlight the possible bottlenecks under a certain set of conditions

Flexibility of the system under changes in demand

Material flow in the distribution facility

ded at each station

Increase throughput, as close as possible to the theoretical maximum per day

Calculate the buffer sizes, if needed, at each station

utilisation

Inter relationship between processes


up and the occurrence of the Soccer World Cup 2010. With

flexible system is of

great importance for the Soccer World Cup. If a team goes through to the next round in the

that team will increase.

the next day, this indicates that some

if the demand can be

The simulation model will be used in studying accurate information and recommendations to

Increase throughput, as close as possible to the theoretical maximum per day

The essence of the parameter will entail the following processes:

� Optimising the processes (

� Material handling, the flow of material through the facility

� Line balancing

� Queuing analysis.

The ultimate goal of the study is to provide Barloworld Logistics wi

the distribution facility’s readiness to cope with the peak demand for the Soccer World Cup. The

study will also indicate what recommendations and changes must be made in order to achieve

that goal.

6

will entail the following processes:

Optimising the processes (E.g. Throughput optimisation)

Material handling, the flow of material through the facility

The ultimate goal of the study is to provide Barloworld Logistics with a model, which will indicate



th a model, which will indicate



1.3 Project Scope

All operation activities regarding the

2). The project is seen as a tactical model for the distribution facility. It will look at the facility and

its outbound operations on a low

its operations. The return station is seen as an operation on its own and the

in the outbound flow of goods and thus will be not be included in the model.

Historical data are used and not the methods of how those data where collected. In essence all

outbound processes are included, from

goods (Figure 2). Figure 3 more clearly illustrates the border of the scope

Figure 3: Scope Boundary

7

ation activities regarding the Outbound Process Flow are included in this project

a tactical model for the distribution facility. It will look at the facility and

low level, taking into account smaller details of every station and

its operations. The return station is seen as an operation on its own and therefore does not fall

flow of goods and thus will be not be included in the model.


are included, from where the orders are made until the despatching of

more clearly illustrates the border of the scope.

are included in this project (Figure

a tactical model for the distribution facility. It will look at the facility and

level, taking into account smaller details of every station and

refore does not fall


until the despatching of

2 Literature Study

2.1 System Analysis Models

2.1.1 Management Decision

It is essential for managers to base their decisions and the extent to which their

made, on the time horizon. The event of the Soccer World Cup takes place in a few months time

and in order to satisfy the demand it is important to make informed decisions. According to

Buzacott & Shanthikumar (1993:5) it is useful to disting

operational decisions, which are defined as follows:

� Strategic – Strategic decisions are those with long time horizon (more than a year or

two) and involve major commitments of the organisation’s resources. These decisions

must be in alignment with the corporate strategy.

� Tactical – Tactical decisions have an interme

months or a lesser scope. These tactical decisions, in turn, become the operating

constraints under which operational planning and control decisions are made.

� Operational – Operational decisions are typically those ma

2.1.2 Nature of Models

According to Buzacott & Shanthikumar (1993:8) a model is based on a representation of the

operation of the system. Buzacott & Shanthikumar (1993:8) states that, the model is ei

mathematical formula or a computer programme that when supplied with the numerical values

of various parameters will make a numerical prediction of the system performance measures.

By changing the values of a parameter the user can gain insight into

performance. Models are intended to support management decisions about the system.

8

System Analysis Models

Management Decision

It is essential for managers to base their decisions and the extent to which their



Buzacott & Shanthikumar (1993:5) it is useful to distinguish between strategic, tactical and

operational decisions, which are defined as follows:

Strategic decisions are those with long time horizon (more than a year or


must be in alignment with the corporate strategy.

Tactical decisions have an intermediate time horizon of between 3 and 18



Operational decisions are typically those made every day or every week.

(Chase, Jacobs & Aquilano 2006:9


operation of the system. Buzacott & Shanthikumar (1993:8) states that, the model is ei



By changing the values of a parameter the user can gain insight into


It is essential for managers to base their decisions and the extent to which their decisions are



uish between strategic, tactical and

Strategic decisions are those with long time horizon (more than a year or


diate time horizon of between 3 and 18



de every day or every week.

Chase, Jacobs & Aquilano 2006:9)


operation of the system. Buzacott & Shanthikumar (1993:8) states that, the model is either a



its influence on


2.2 Basic Approach to Modelling

The process of modelling involves the following steps:

� Identify the issues to be addressed

The needs of management must be e

The decisions that the model is required to support have to be identified.

� Learn about the system

Buzacott & Shanthikumar

the machines, material handling and the data collection and control systems. Determine

the characteristics of the job and the target volumes, quality and cost

review any existing models to identify their capabilities, shortcomings and the level of

familiarity of managers with the use of

design or operation.

� Choose a modelling approach

“Various types of modelling

models through computer simulations

by the time and cost budget

level of user interface should also be considered when selecting the

� Develop and test the model

During this step, data on the parameters of the model should be acquired. Decisions on

the level of detail, accuracy and the method of collecting data should be made, since it

will influence the correctness of the system, Buzacott & Shanthikumar (1993:8).

9

Modelling

involves the following steps:

Identify the issues to be addressed

The needs of management must be established before clarifying the major concerns.


Learn about the system

Buzacott & Shanthikumar (1993:8) said, “Identify the elements of the system, such as

handling and the data collection and control systems. Determine

the characteristics of the job and the target volumes, quality and cost


f managers with the use of modelling approaches to help in the system

approach

modelling approaches can be used, ranging from formal mathematical

models through computer simulations. The choice of modelling approach is determined

by the time and cost budget” said by said by Buzacott & Shanthikumar (1993:8)

level of user interface should also be considered when selecting the modelling

Develop and test the model

on the parameters of the model should be acquired. Decisions on


e the correctness of the system, Buzacott & Shanthikumar (1993:8).

stablished before clarifying the major concerns.


the elements of the system, such as

handling and the data collection and control systems. Determine

the characteristics of the job and the target volumes, quality and cost”. It is useful to


approaches to help in the system

approaches can be used, ranging from formal mathematical

approach is determined

said by said by Buzacott & Shanthikumar (1993:8). The

modelling approach.

on the parameters of the model should be acquired. Decisions on


e the correctness of the system, Buzacott & Shanthikumar (1993:8).

� Verify and validate the model

The model should be checked to see that it is a reasonably correct representation of

reality; a base case should be established. This includes verification and validation.

• “Verification is to ensure the model results are correct derived from the

assumptions made in developing the model

(1993:8).

• “Validation is the process of ensuring that the assumptions of the model and the

results are an accurate representation of the real system

Shanthikumar (1993:8).

� Develop a model interface for the user

Buzacott & Shanthikumar (1993:8) said “

decisions it has to be provided with some interface so that it can be used by

management”. This requires the modeller to eit

support system or to present the model and its implications in a way that management

can understand. The user has to be convinced that the model is adequate for the

decision that the model is required to support.

� Experiment with the model

“This requires exploring the impact of changes in model parameters and developing

understanding of the factors influencing performance of the system so that the manager

can be confident in the decisions made using the model


10

the model



Verification is to ensure the model results are correct derived from the

assumptions made in developing the model” said by Buzacott & Shanthikumar

Validation is the process of ensuring that the assumptions of the model and the

results are an accurate representation of the real system” said by Buzacott &

r (1993:8).

Develop a model interface for the user

Buzacott & Shanthikumar (1993:8) said “If the model is to be of value in making


. This requires the modeller to either incorporate the model into a decision



decision that the model is required to support.

Experiment with the model

This requires exploring the impact of changes in model parameters and developing


can be confident in the decisions made using the model” said



Verification is to ensure the model results are correct derived from the

” said by Buzacott & Shanthikumar

Validation is the process of ensuring that the assumptions of the model and the

” said by Buzacott &

If the model is to be of value in making


her incorporate the model into a decision



This requires exploring the impact of changes in model parameters and developing


” said by Buzacott &

� Present the Results

Using the model, the manager should implement the recommended course of action. It

is important that the users are informed about the accuracy of the results which will be

generated by the model


Figure 4: Basic Approach to Modelling

11



in order to make correct and informed decisions, Buzacott &

Modelling



correct and informed decisions, Buzacott &

2.3 Types of Models

For discrete part manufacturing there are three types of models in common use:

2.3.1 Physical Models

(Buzacott & Shanthikumar 1993:11) said, “

another physical system, in which parts move from one machine to another and the machines

perform processing operations on the parts. The major difference

the model uses a different dimensional scale, so a large system will occupy a large amount

space. Physical models are excellent as a means of educating production management and

workers about the control of the system, but the

run behaviour of the system as it is difficult to represent the statistical properties of events such

as machine failures”.

2.3.2 Computer Simulation Models

Kelton, Sadowski & Sturrock (2007:5) said that, “

studying a wide variety of models of real

designed to imitate the system’s operations or characteristics, often over time. From a practical

viewpoint, simulation is the process of designing and creating a computerised model of a real or

proposed system for the purpose of conducting numerical experiments to give us a better

understanding of the behaviour of that system for a given set of conditions

can become quite complicated, if needed to represent the system faithfully and you can

simulation analysis.

“The probabilistic nature of many events, such as machine failure, can be represented by

sampling from a distribution representing the patte

represent the typical behaviour of the system it is necessary to run the simulation model for a

sufficient long time, so that all events can occur a reasonable number of times

Buzacott & Shanthikumar (1993:1

graphic display to demonstrate the movement of jobs and material handling

12


(Buzacott & Shanthikumar 1993:11) said, “Physical models represent the real system by


perform processing operations on the parts. The major difference from the real system is that



workers about the control of the system, but they do not lend themselves to assessing the long


Computer Simulation Models

Kelton, Sadowski & Sturrock (2007:5) said that, “Computer simulation refers to methods for

studying a wide variety of models of real-world systems by numerical evaluation using software


process of designing and creating a computerised model of a real or


understanding of the behaviour of that system for a given set of conditions”. Simulation models

, if needed to represent the system faithfully and you can

The probabilistic nature of many events, such as machine failure, can be represented by

sampling from a distribution representing the pattern of occurrence of the event. Thus to


sufficient long time, so that all events can occur a reasonable number of times

Buzacott & Shanthikumar (1993:11). Simulation models can be provided with an interactive

graphic display to demonstrate the movement of jobs and material handling devices. This can


Physical models represent the real system by


from the real system is that



y do not lend themselves to assessing the long


simulation refers to methods for

world systems by numerical evaluation using software


process of designing and creating a computerised model of a real or


. Simulation models

, if needed to represent the system faithfully and you can still do a

The probabilistic nature of many events, such as machine failure, can be represented by

rn of occurrence of the event. Thus to


sufficient long time, so that all events can occur a reasonable number of times”, said by

. Simulation models can be provided with an interactive

devices. This can

be of great value in communicating the assumptions of the model to the managers (Buzacott &

Shanthikumar 1993:11).

2.3.3 Analytical Models

Analytical models describe the system using mathematical or symbolic relationships. If the

model is simple, traditional mathematical tools like queuing theory, differential

or linear programming can be used (K

Shanthikumar 1993:12 said that, “

algorithm or computational procedure by which the performance measures of the system can be

calculated. Analytical models can also be used to demonstrate properties of various operating

rules and control strategies”.

2.3.4 Analytical versus Simulation Models

Simulation models can be developed in several different ways. One is to try and include all the

required detail of the system from the beginning. This tends to result in long development and

debugging time, and usually makes the resultant code unreadable to all but the developer.

Another approach is to develop models of the system components and after each component

model has been debugged, tested and validated, link the components two at a time, test and

validate, then three at a time, test and validate and so on. Sometimes there are already existing

simulation models of various sub

overall system model but often it is possible to devise ways of summarising the performance of

the sub-system model and replacing it by a simple sub

approach is called the micro-macro approach and is useful

systems.

Simulation models are usually written in a simulation language such as GPSS, SLAM, SIMAN,

SIMULA or SMALLTALK and only experts in the simulation language can read the code.

13



model is simple, traditional mathematical tools like queuing theory, differential-equation methods

or linear programming can be used (Kelton, Sadowski & Sturrock 2007:5).

Shanthikumar 1993:12 said that, “These are then used to derive a formula or to define an


s can also be used to demonstrate properties of various operating

Analytical versus Simulation Models


tem from the beginning. This tends to result in long development and



been debugged, tested and validated, link the components two at a time, test and


simulation models of various sub-systems. The problem is that these can be too detailed for the


system model and replacing it by a simple sub-system component model. This

macro approach and is useful in developing models of very large





equation methods

elton, Sadowski & Sturrock 2007:5). Buzacott &

These are then used to derive a formula or to define an


s can also be used to demonstrate properties of various operating


tem from the beginning. This tends to result in long development and



been debugged, tested and validated, link the components two at a time, test and


oo detailed for the


system component model. This

in developing models of very large



Analytical models typically determine

whereas simulation models determine a time

statistical test is required to determine whether the sample is consistent with the analytical

results. Analytical models either give a formula or define a computational procedure. If they give

a computational procedure then it is necessary to consider the computational complexity of the

procedure. It is by no means uncommon for the computational requirements to increase

exponentially. This often limits the size of the system and the range of parameter values for

which the procedure can be used (Buzacott & Shanthikumar 1993:12).

2.4 Why Model?

Models can be developed for a variety of reasons, in particular:

� Understanding

“The model is used to explain why and how. The model is intended to help develop

insight. It also indicates the direction of influence of some variable on performance

by Buzacott & Shanthikumar

� Learning

“As well as providing insight a model may

about the factors that determine performance

(1993:12).

� Improvement

“The model is used to improve system design and operation. Changes in parameters

and rules can be explored, an

identified” said by Buzacott & Shanthikumar

14

Analytical models typically determine the expected value of some performance criterion,

whereas simulation models determine a time-limited sample of performance. Thus some


either give a formula or define a computational procedure. If they give



nentially. This often limits the size of the system and the range of parameter values for

which the procedure can be used (Buzacott & Shanthikumar 1993:12).

Models can be developed for a variety of reasons, in particular:

el is used to explain why and how. The model is intended to help develop

insight. It also indicates the direction of influence of some variable on performance

Buzacott & Shanthikumar (1993:12).

As well as providing insight a model may be intended to teach managers or workers

about the factors that determine performance” is said by Buzacott & Shanthikumar

The model is used to improve system design and operation. Changes in parameters

and rules can be explored, and factors critical for achieving performance targets can be

” said by Buzacott & Shanthikumar (1993:12).

the expected value of some performance criterion,

limited sample of performance. Thus some


either give a formula or define a computational procedure. If they give



nentially. This often limits the size of the system and the range of parameter values for

el is used to explain why and how. The model is intended to help develop

insight. It also indicates the direction of influence of some variable on performance”, said

be intended to teach managers or workers

” is said by Buzacott & Shanthikumar

The model is used to improve system design and operation. Changes in parameters

d factors critical for achieving performance targets can be

Understanding

Decision

making

� Optimisation

“The model is to be used to aid decisions about either the design or operation of the

system. The model has to be able to dif

and project their impact over time

Figure 5: Areas influenced by modelling

2.5 Simulation Modelling

Simulation may be defined as a technique that imitates the operation of the real

as it evolves over a period of time. There are two types of simulation models: Static and

dynamic. A static simulation model represents a system at a particular point in time. A dy

simulation model represents a system as it evolves over time.

15

Understanding Learning

Decision

makingImprovement

Optimisation

Model

The model is to be used to aid decisions about either the design or operation of the

system. The model has to be able to differentiate the effect of different courses of action

and project their impact over time”, is said by Buzacott & Shanthikumar 1993:12

modelling

Modelling

defined as a technique that imitates the operation of the real


dynamic. A static simulation model represents a system at a particular point in time. A dy

simulation model represents a system as it evolves over time.

The model is to be used to aid decisions about either the design or operation of the

ferentiate the effect of different courses of action

”, is said by Buzacott & Shanthikumar 1993:12.

defined as a technique that imitates the operation of the real-world system


dynamic. A static simulation model represents a system at a particular point in time. A dynamic

Winston (2004:443) said that, “

deterministic simulation contains no random variables, whereas a stochastic simulation cont

one or more random variables”. Finally, simulations may be represented by either discrete or

continuous models. A discrete simulation is one in which the state variable change only at

discrete points in time. In a continuous simulation, the state vari

time.

Simulation moves the real world to a virtual world of models, performing experiments with the

model of the system in a risk-free environment and map the best possible solution back to the

real world see Figure 6.


16

Winston (2004:443) said that, “Simulations can be divided into deterministic or stochastic. A

deterministic simulation contains no random variables, whereas a stochastic simulation cont

. Finally, simulations may be represented by either discrete or


discrete points in time. In a continuous simulation, the state variables change continuously over


free environment and map the best possible solution back to the

: Illustrating the mapping of simulation between the real world and models

Simulations can be divided into deterministic or stochastic. A

deterministic simulation contains no random variables, whereas a stochastic simulation contains

. Finally, simulations may be represented by either discrete or


ables change continuously over


free environment and map the best possible solution back to the

: Illustrating the mapping of simulation between the real world and models

2.5.1 Simulation Language

Computer programming is the basis of simulation

for a complex problem is a difficult and

simplify computer programming. The most widespread and readily available simula

languages include GPPS, GASP, ACSL, SIMSCRIPT, SLAM

2004:440).

2.5.2 High-Level Simulators

High-Level simulators typically operate by intuitive graphical user interface, menus and dialogs.

Selecting from available simulation

along with a dynamic graphical animation of system components as they move around and

change (Kelton, Sadowski & Sturrock 2007:5).

2.5.3 Simulation Modelling

Currently in the markets there are numerous si

own capabilities. Each tool has a different approach to simulation and their success factors vary

in more ways than one. Because of the great number of tools available only a few, which is

most applicable to this project will be analysed further.

� Arena®

“Rockwell Arena is simulation and automation software acquired by Rockwell Automation

almost a decade ago. It uses the SIMAN processor and simulation language. It is currently

in version 12.0”.

[online]http://en.wikipedia.org/wiki/Rockwell_Arena

17

Simulation Language

programming is the basis of simulation. The study and writing of the computer code

for a complex problem is a difficult and laborious task. Special purpose simulation languages

simplify computer programming. The most widespread and readily available simula

S, GASP, ACSL, SIMSCRIPT, SLAM, SIMAN and many more

Level Simulators

Level simulators typically operate by intuitive graphical user interface, menus and dialogs.

Selecting from available simulation-modelling constructs connects them and run the model


change (Kelton, Sadowski & Sturrock 2007:5).

Modelling Tools

Currently in the markets there are numerous simulating tools available, each of which has its



this project will be analysed further.

Rockwell Arena is simulation and automation software acquired by Rockwell Automation


(Arena Rockwell Software

http://en.wikipedia.org/wiki/Rockwell_Arena,accessed on 14May 2009)

the computer code

task. Special purpose simulation languages

simplify computer programming. The most widespread and readily available simulation

and many more (Winston

Level simulators typically operate by intuitive graphical user interface, menus and dialogs.

constructs connects them and run the model


mulating tools available, each of which has its



Rockwell Arena is simulation and automation software acquired by Rockwell Automation


Software – wikipedia.

accessed on 14May 2009)

Arena® combines the ease of use found in high

simulation languages and accommodates general

Microsoft® Visual Basic® programming system and even C.

Arena maintains its modelling

templates of graphical simulation

be used at any time and access to simulation language flexibility is unlimited. The Arena®

software is completely hierarchical allowing access to a whole set of simulation

constructs and capabilities, such as the SIMAN constructs together with higher

modules from other templates.

Arena® is used most effectively

involving highly sensitive changes relat

logistics, distribution, warehousing and service systems

Applications include:

• “Analysis of Six Sigma implementation, KanBan/Pull or Push systems

• “Detail analysis of warehousing, logistics or

applications”.

• “Modelling complex customer

• “Detail analysis of complex manufacturing processes that include material

intensive operations”.

http://www.arenasimulation.com/products/professional_edition.asp, accessed

18

Arena® combines the ease of use found in high-level simulators with the flexibility of

n languages and accommodates general-purpose procedural languages like the

Microsoft® Visual Basic® programming system and even C.

modelling flexibility by allowing the user to interchange between

templates of graphical simulation modelling and analysis modules. Low-level modules can


software is completely hierarchical allowing access to a whole set of simulation

es, such as the SIMAN constructs together with higher

modules from other templates.

(Kelton, Sadowski & Sturrock 2007:10)

Arena® is used most effectively “when analysing complex, medium to large

involving highly sensitive changes related to supply chain, manufacturing, processes,

logistics, distribution, warehousing and service systems”.

Analysis of Six Sigma implementation, KanBan/Pull or Push systems

Detail analysis of warehousing, logistics or transportation, military and mining

complex customer-handling activities”.

Detail analysis of complex manufacturing processes that include material

”.

(Arena® Professional Edition. [online]

http://www.arenasimulation.com/products/professional_edition.asp, accessed

level simulators with the flexibility of

purpose procedural languages like the

flexibility by allowing the user to interchange between

level modules can


software is completely hierarchical allowing access to a whole set of simulation modelling

es, such as the SIMAN constructs together with higher-level


when analysing complex, medium to large-scale projects

ed to supply chain, manufacturing, processes,

Analysis of Six Sigma implementation, KanBan/Pull or Push systems”.

transportation, military and mining

Detail analysis of complex manufacturing processes that include material-handling

(Arena® Professional Edition. [online]

http://www.arenasimulation.com/products/professional_edition.asp, accessed on 14May

2009)

Arena® exploits two Microsoft Windows® technologies that are designed to en

integration of desktop applications. The first, ActiveX® Automation, allows applications to

control each other and themselves via a programming interface. Programming languages

like C++, Visual Basic or Java is used to create the programme that c

The second technology exploited by Arena® for application integration is Visual Basic

(VBA). VBA is the same language that works with Microsoft Office, AutoCAD and Arena®.

These two technologies work together to allow Arena® to i

that support ActiveX Automation. It is possible to write Visual Basic code directly in Arena®

that automates other programmes, such as Excel, AutoCAD or Visio.

It has been noted that creating a simulation can require more time at the beginning of a

project, but quicker installations and product optimizations can reduce overall project time.

19

Arena® exploits two Microsoft Windows® technologies that are designed to en



like C++, Visual Basic or Java is used to create the programme that controls the application.



These two technologies work together to allow Arena® to integrate with other programmes


that automates other programmes, such as Excel, AutoCAD or Visio.


ating a simulation can require more time at the beginning of a


Arena® exploits two Microsoft Windows® technologies that are designed to enhance the



ontrols the application.



ntegrate with other programmes



ating a simulation can require more time at the beginning of a


� ProModel

“ProModel is discrete event simulation software, used for evaluating, planning

manufacturing, warehousing, logistics and other operational and strategic situations.

This easy-to-use software allows you to build computer models of your situation and

experiment with scenarios to find the best solution. The software include

graphical reports, providing powerful tools for analysing and optimisation

• Develop “what if” scenarios quickly

• Direct data import and export with Microsoft® Excel®

• Capture system randomness and variability by utilising over 20 statistical

types or directly import your own data.

• Distribute models to other divisions and department with run

• Automatic creation of output reports

ProModel is an easy-to-use and easy

http://www.promodel.com/products/promodel/features.asp

� ACSL

“The Advanced Continuous Simulation Language, or ACSL (pronounced “axle’), is a computer language designed for continuous system described by time

http://en.wikipedia.org/wiki/advanced_continuous_simulation_language

20

ProModel is discrete event simulation software, used for evaluating, planning


use software allows you to build computer models of your situation and

experiment with scenarios to find the best solution. The software include

graphical reports, providing powerful tools for analysing and optimisation”.

Develop “what if” scenarios quickly

Direct data import and export with Microsoft® Excel®

Capture system randomness and variability by utilising over 20 statistical

types or directly import your own data.

Distribute models to other divisions and department with run-time licensing

Automatic creation of output reports

use and easy-to-implement.

(ProModel Optimisation Software. [onli

http://www.promodel.com/products/promodel/features.asp, accessed on 13May 2009)

The Advanced Continuous Simulation Language, or ACSL (pronounced “axle’), is a designed for modelling and evaluating the performance of

continuous system described by time-dependent, nonlinear differential equations

(Acsl-wikipedia. [online]

http://en.wikipedia.org/wiki/advanced_continuous_simulation_language, accessed on 11May

ProModel is discrete event simulation software, used for evaluating, planning or designing


use software allows you to build computer models of your situation and

experiment with scenarios to find the best solution. The software includes animation and

Capture system randomness and variability by utilising over 20 statistical distribution

time licensing

(ProModel Optimisation Software. [online]

, accessed on 13May 2009)

The Advanced Continuous Simulation Language, or ACSL (pronounced “axle’), is a and evaluating the performance of

dependent, nonlinear differential equations”.

wikipedia. [online]

, accessed on 11May

2009)

ACSLX – “ACSLX is the latest software package by ACSL. ACSLX is a

execution and analysis environment for continuous dynamic systems and processes.

Simple to learn and easy to use, ACSLXtream provid

users at all level, is versatile and powerful enough to address the most challenging

simulation problems”.

The ready-to-use code blocks incorporated in ACSLX enable quick model assembly,

while powerful analysis capabilities provide quick and accurate results. Models are

specified using either familiar block diagram notation or code

CSSL (Continuous System Simulation Language)

representations are translated into C or FORTRAN language source code for

compilation and execution, ensuring the fastest possible performance and support for

extremely large or complex

(ACSLX. [online]

The Arena® package developed by Rockwell Software will be selected, investigation shows that

it is the best software for the project and

the University of Pretoria without additional expenses.

21

ACSLX is the latest software package by ACSL. ACSLX is a


imple to learn and easy to use, ACSLXtream provides an intuitive environment for


use code blocks incorporated in ACSLX enable quick model assembly,


specified using either familiar block diagram notation or code-based descriptions using

(Continuous System Simulation Language) modelling language. All model



extremely large or complex models.

(ACSLX. [online] http://www.acslsim.com/products, accessed on 13May 2009)


roject and that a full version of this package can be provided by

the University of Pretoria without additional expenses.

ACSLX is the latest software package by ACSL. ACSLX is a modelling,


es an intuitive environment for


use code blocks incorporated in ACSLX enable quick model assembly,


based descriptions using

language. All model



, accessed on 13May 2009)


that a full version of this package can be provided by

3 Conceptual Design of Model

3.1 Structuring the Model

A basic approach will be followed to construct the simulation model. The major

simulation study are illustrated in

simulation study, simplifies the modelling

model are made as the system is under examination. The goal of the initial model is to create a

simulation model that represents the current operation of the distribution facility.

According to Chase, Jacobs and Aquilano (2006:703) a continuous model is based on

mathematical equations with values for all points in time.

Discrete models are a process, in which changes to the state of the system occur at isolated

points in time (Kelton, Sadowski & Sturr

at Nike is a discrete simulation, events occur at specific points in time. Goods arriving at the

distribution facility are an example of a discrete event.

In a system, an object of interest is call

attributes (Winston 2004:443). The

of the attributes (one entity may have more than one attribute) attaches to that entity is the

of business units ordered by the customer,

The simulation model is also a queuing system,

station and waiting to be sent through the system.

Aquilano (2006:693).

22

Conceptual Design of Model

Structuring the Model

A basic approach will be followed to construct the simulation model. The major

simulation study are illustrated in Figure 7. This flow chart view of the major phases in

modelling process significantly, especially when changes to the



Jacobs and Aquilano (2006:703) a continuous model is based on

mathematical equations with values for all points in time.

a process, in which changes to the state of the system occur at isolated

points in time (Kelton, Sadowski & Sturrock 2007:465). The modelling of the distribution facility


distribution facility are an example of a discrete event.

In a system, an object of interest is called an entity and any properties of an entity are called

attributes (Winston 2004:443). The entity of this simulation is an order from a customer

of the attributes (one entity may have more than one attribute) attaches to that entity is the

business units ordered by the customer, i.e. footwear, apparel or equipment.

The simulation model is also a queuing system, with many orders or entities arriving at each

station and waiting to be sent through the system. Figure 7 is obtained from Chase, Jacobs and

A basic approach will be followed to construct the simulation model. The major phases in the

. This flow chart view of the major phases in

process significantly, especially when changes to the



Jacobs and Aquilano (2006:703) a continuous model is based on

a process, in which changes to the state of the system occur at isolated

of the distribution facility


ed an entity and any properties of an entity are called

entity of this simulation is an order from a customer and one

of the attributes (one entity may have more than one attribute) attaches to that entity is the type

or entities arriving at each

is obtained from Chase, Jacobs and

Figure 7: Major Phases in Simulation Study

23

: Major Phases in Simulation Study

Chase, Jacobs and Aquilano (2006:291) states a queuing system consists essentially of three

major components:

1. The source population and the ways entities arrive at the system.

2. The service system.

3. The condition of the entity exiting the system.

3.2 Conceptual Inputs and Outputs of the

The input analysis specifies the model parameters and distributions. The logic aspect of the

model, such as the entity and resource characteristics, the path an entity follows through the

system and any other activities is called structural modelling

2007:172).

The structural modelling of a system is an arrangement of the fundamental logic that describes

the system. The mathematical input that describes the logic is called quantitative

Quantitative modelling specifies the numerical nature of the resources and entities included and

not only, the distributions and model parameters modelling (Kelton, Sadowski & Sturrock

2007:172).

The inputs to the queuing model are the arrival process, the process time of each s

number of entities and the number of lines.

The structural modelling of a simple flow chart for an entity arrival is illustrated by Winston

(2004:405) in Figure 8.

24

(2006:291) states a queuing system consists essentially of three

The source population and the ways entities arrive at the system.

The condition of the entity exiting the system.

Conceptual Inputs and Outputs of the Model



system and any other activities is called structural modelling (Kelton, Sadowski & Sturrock

of a system is an arrangement of the fundamental logic that describes

the system. The mathematical input that describes the logic is called quantitative

pecifies the numerical nature of the resources and entities included and


The inputs to the queuing model are the arrival process, the process time of each s

number of entities and the number of lines.

of a simple flow chart for an entity arrival is illustrated by Winston

(2006:291) states a queuing system consists essentially of three



(Kelton, Sadowski & Sturrock

of a system is an arrangement of the fundamental logic that describes

the system. The mathematical input that describes the logic is called quantitative modelling.

pecifies the numerical nature of the resources and entities included and


The inputs to the queuing model are the arrival process, the process time of each station, the

of a simple flow chart for an entity arrival is illustrated by Winston

Figure 8: Flow chart of an Arrival

The values of the inputs to the simulation are determined by using Microsoft Excel®. Excel®

spreadsheets are used as a user

which will be used in the simulation model.

By linking these spreadsheets with the model, the input values can be altered for different

scenarios without having to change the values in the simulation model itself. All the

spreadsheets are linked with each other, providing a relationship between interrelated cells.

This makes it easier for a person with no knowledge of how Arena® operates to run different

scenarios and draw data from the model.

25

values of the inputs to the simulation are determined by using Microsoft Excel®. Excel®

spreadsheets are used as a user-interface tool to assign values to attributes and variables,

which will be used in the simulation model.



eets are linked with each other, providing a relationship between interrelated cells.


scenarios and draw data from the model.

values of the inputs to the simulation are determined by using Microsoft Excel®. Excel®

ttributes and variables,



eets are linked with each other, providing a relationship between interrelated cells.


The conceptual outputs of the queuing system are typified by a specific probability distribution,

which governs a customer’s service time. The random variable input data generates random

variability in the output data. The most important output statistic of a simulation model will

include the following performance measures (Mehrotra & Fama

• Abandonment Statistics

• Queue Statistics

• Volume Statistics

The system will be measured by several KPI’s (Key Performance Indicators)

used to measure the performance of

important KPI’s will include the following

1. Resource Utilisation (Picker Utilisation and Packer Utilisation)

time that a resource is truly being used in relation to the time that it is available.

2. Throughput Time – It includes the time that the unit is being worked on together with

the time that the unit spends waiting in a queue

through the system).

3. Time in Process – The time a unit takes to be completed in a specific process e.g. the

time it takes a unit to pass through the picking phase.

26

queuing system are typified by a specific probability distribution,



ude the following performance measures (Mehrotra & Fama 2003:137).

he system will be measured by several KPI’s (Key Performance Indicators). These KPI’s are

measure the performance of the different scenarios against each other.

will include the following as stated by Chase, Jacobs & Aquilano

(Picker Utilisation and Packer Utilisation) – It is the ratio of the

e is truly being used in relation to the time that it is available.

It includes the time that the unit is being worked on together with

the time that the unit spends waiting in a queue (The average time a unit takes to move

The time a unit takes to be completed in a specific process e.g. the

time it takes a unit to pass through the picking phase.

queuing system are typified by a specific probability distribution,



. These KPI’s are

the different scenarios against each other. The more

Chase, Jacobs & Aquilano (2006:164).

It is the ratio of the

e is truly being used in relation to the time that it is available.

It includes the time that the unit is being worked on together with

(The average time a unit takes to move

The time a unit takes to be completed in a specific process e.g. the

4 Data Collection and Analysis

One of the very early steps in planning a simulation project should be to identify what data is

needed to support the model. The availability and quality of data can influence the

approach taken and the level of detail captured in the model

2007:174).

Accourding to Kelton, Sadowski & Sturrock (2007:175) a benefit of developing an early

understanding of the quality of your input data is that it can help you decide how much detail to

incorporate in the model logic. The

study are only as reliable as the model and its inputs.

Input Data Analyzer is included in the Arena® software, it automatically fits existing data values

to the best statistical probability

data and provides an estimate of the parameter values and an expression of the model (Kelton,

Sadowski & Sturrock 2007:176).

4.1 Input Data and Analysis

Data will be received by intensive time

retrieved from the Microsoft Access® data base. It is then transferred to Microsoft Ex

is more accessible and the data is analysed.

4.1.1 Order Information

Each order generated by a customer has cert

ordered, the quantity of each item and the delivery date. Each order is then classified into three

different channels, the IDP (Integrated Delivery Plan), Non IDP and the Priorities. The channels

differ in the delivery dates, how urgent a certain order is. The IDP has a fixed despatch date and

planning is done around that date. The Non IDP is the most flexible channel it has a shipping

27

Data Collection and Analysis


needed to support the model. The availability and quality of data can influence the

approach taken and the level of detail captured in the model (Kelton, Sadowski & Sturrock

Kelton, Sadowski & Sturrock (2007:175) a benefit of developing an early


incorporate in the model logic. The results and recommendations present, from the simulation

study are only as reliable as the model and its inputs.


to the best statistical probability distributions. The Input Data Analyzer fits a distribution to the


Sadowski & Sturrock 2007:176).

and Analysis

Data will be received by intensive time studies on the relevant stations. The input data is

retrieved from the Microsoft Access® data base. It is then transferred to Microsoft Ex

is more accessible and the data is analysed.

by a customer has certain data attached to it. The types



n the delivery dates, how urgent a certain order is. The IDP has a fixed despatch date and



needed to support the model. The availability and quality of data can influence the modelling

on, Sadowski & Sturrock

Kelton, Sadowski & Sturrock (2007:175) a benefit of developing an early


from the simulation


distributions. The Input Data Analyzer fits a distribution to the


studies on the relevant stations. The input data is

retrieved from the Microsoft Access® data base. It is then transferred to Microsoft Excel®, which

n data attached to it. The types of goods that is



n the delivery dates, how urgent a certain order is. The IDP has a fixed despatch date and


window of delivery date plus 30 days. T

in today it must be despatched the next day, making it the least flexible channel.

the most important channel for the world cup season.

4.1.2 Transport Time of a Picker per Order

The Transport time of a picker per ord

from the one bin location to the next in order to pick the units

consuming as the order could contain any of the 30,000+ products Nike offer.

Each unit has a specified location that is entered into the system. The levels of the pick face is

divided into blocks, level A is divided into 3x9 blocks and levels B,C and D is divided into 3x12

blocks. The blocks are defined by an

and the distances from one block to the next. The locations of the units are then incorporated

into one of these blocks, thus each un

are linked via there reference to a blo

The time it takes a picker to walk from one block to the next is calculated through time studies

(see Figure 9). Level A differs from levels B

locations on these levels.

The time to walk from one block to the next or to any other block is calculated in Microsoft

Access®. A formula is used where the times are multiplied if the picker either walks in

direction or in the y-direction. The formula differs from level A to the others

formula is multiplied with. E.g. level A is multiplied in the x

multiplied with 14 (see Figure 9 for the values).

28

livery date plus 30 days. The Priorities is the most urgent orders, if an order comes

ust be despatched the next day, making it the least flexible channel.

the most important channel for the world cup season.

Transport Time of a Picker per Order

The Transport time of a picker per order is the time a picker takes to walk on a specific level

from the one bin location to the next in order to pick the units to fulfil the order

consuming as the order could contain any of the 30,000+ products Nike offer.

specified location that is entered into the system. The levels of the pick face is

, level A is divided into 3x9 blocks and levels B,C and D is divided into 3x12

ks. The blocks are defined by an X-Y co-ordination in order to calculate the size of a block


into one of these blocks, thus each unit belongs to a specific block on a specific level.

to a block.


). Level A differs from levels B, C and D because of the greater number of


Access®. A formula is used where the times are multiplied if the picker either walks in

direction. The formula differs from level A to the others by;

formula is multiplied with. E.g. level A is multiplied in the x-direction with 11 and levels B, C, D is

for the values).

nt orders, if an order comes

ust be despatched the next day, making it the least flexible channel. This channel is

er is the time a picker takes to walk on a specific level

order. This is time

specified location that is entered into the system. The levels of the pick face is

, level A is divided into 3x9 blocks and levels B,C and D is divided into 3x12

e the size of a block


it belongs to a specific block on a specific level. The units


and D because of the greater number of storage


Access®. A formula is used where the times are multiplied if the picker either walks in the x-

by; the number the

direction with 11 and levels B, C, D is

Figure 9: Walking time from one block to next

The travel times are then exported to Microsoft Excel® where the data are pivoted into a table

for travel times of level A and a separate table for levels B,

15 and Table 16). If picking at a certain level does occur, the transport time is calculated in

Microsoft Excel® using the Index and Match formulas. The formula is an “if” statement, it looks

on what level the picking must occur and uses the level A table or the level B, C, D

“if” statement the Index function is used marking the table that needs to be used. In combination

with Index function the Match function is used in order the match

starting and ending locations), the picker must

one block to the next block where the picker must pick next.

After the travel times between the blocks, in order to pick units,

time to the conveyor is calculated

29

: Walking time from one block to next

then exported to Microsoft Excel® where the data are pivoted into a table

for travel times of level A and a separate table for levels B, C and D (see Appendix

picking at a certain level does occur, the transport time is calculated in


on what level the picking must occur and uses the level A table or the level B, C, D


with Index function the Match function is used in order the match, from where and

the picker must walk. This value is the transport time from the

one block to the next block where the picker must pick next.

between the blocks, in order to pick units, are calculated

calculated (after all picks is completed and the units must be taken to the

then exported to Microsoft Excel® where the data are pivoted into a table

ppendix 10.1, Table

picking at a certain level does occur, the transport time is calculated in


on what level the picking must occur and uses the level A table or the level B, C, D table. In the


where and to where (the

This value is the transport time from the

are calculated then the travel

ks is completed and the units must be taken to the

conveyor). There are three blocks in the x

irrelative in what y-block the picker is, is the block number in the x

E.g. If the picker stands in the second block in the x

conveyor is 2*7.5 = 15sec. This is calculated for every order on a specific floor.

The total transport time for a picker is the sum of the transport times. These

expanded into their different levels (A, B, C, D) by placing them into a matrix form (see

Appendix 10.1, Table 17).

4.1.3 Quantity of a Business Unit Picked

An order contains the quantity of certain

Access® database then finds the loc

quantity units, type of business unit (footwear or other) and the level it is situated in. A matrix

table is then compiled of the information in Microsoft Excel® from the data in the Microsoft

Access® database. The matrix table

quantity that must be picked on a level (s

computation between Microsoft Access® and Excel® is done automatically

4.1.4 Number to be batched

An order can have units that need

need to be picked on different levels,

different levels. This split in the order must be batched toge

boxes. The number to be batched calculates

be batched must be equal to the number of splits

not be despatched until the number of splits are all together

to assure that all orders are complete when

30

. There are three blocks in the x-direction, the time it takes to walk to the conveyor

block the picker is, is the block number in the x-direction multiplied with 7.5.

If the picker stands in the second block in the x-direction then the transport time to the

conveyor is 2*7.5 = 15sec. This is calculated for every order on a specific floor.

The total transport time for a picker is the sum of the transport times. These


Quantity of a Business Unit Picked

An order contains the quantity of certain products and the type of business units. The Microsoft

Access® database then finds the locations of the products that are ordered. It captures the


of the information in Microsoft Excel® from the data in the Microsoft

The matrix table states the order number; the type of business units and the

that must be picked on a level (see Appendix 10.1, Table 18). Each order differs and all

computation between Microsoft Access® and Excel® is done automatically for each order

hed

units that need to be picked on different levels. If an order has

need to be picked on different levels, it is then split by the units that need

different levels. This split in the order must be batched together when the order is packed into

number to be batched calculates how many splits an order has and the number to

be batched must be equal to the number of splits, this is illustrated in Figure 10

not be despatched until the number of splits are all together (see Appendix 10.1

orders are complete when despatched.

to walk to the conveyor

direction multiplied with 7.5.

direction then the transport time to the

The total transport time for a picker is the sum of the transport times. These times are then


and the type of business units. The Microsoft

that are ordered. It captures the


of the information in Microsoft Excel® from the data in the Microsoft

business units and the

). Each order differs and all

for each order.

an order has units that

to be picked at

ther when the order is packed into

and the number to

10. The order will

10.1, Table 19). It is

Figure 10: Illustration of the split of an Order

4.1.5 Distance Conveyor travel in Picking

On the side of every level is a conveyor (see

order, the tote (picked units) is then placed

packing. The conveyor moves from level to level

Each order ends on a different place on a level. The picker then walks straight

picked locations to the conveyor. The

has an influence on the distance

on the time it takes a tote to travel to packing.

If a tote is placed at position A,

point A to the end of level A, shown as a blue arrow

then the tote must move the distance on level D and the entire distance on level C

green arrow. Thus the tote at point B will travel longer to get to packing than point A.

31

split of an Order

Distance Conveyor travel in Picking

On the side of every level is a conveyor (see Figure 9), when a picker is finished picking

is then placed on the conveyor. The conveyor conveys the totes to

The conveyor moves from level to level as illustrated in Figure 11 by the red arrows

Each order ends on a different place on a level. The picker then walks straight

to the conveyor. The location of where the totes are placed on the conveyor

has an influence on the distance that the tote must travel to packing; this has a direct influence

on the time it takes a tote to travel to packing.

in Figure 11, then the tote only has to travel the distance from

, shown as a blue arrow. If the tote is placed at point B in

then the tote must move the distance on level D and the entire distance on level C

at point B will travel longer to get to packing than point A.

a picker is finished picking an

on the conveyor. The conveyor conveys the totes to

by the red arrows.

Each order ends on a different place on a level. The picker then walks straight from the last

placed on the conveyor

this has a direct influence

, then the tote only has to travel the distance from

ed at point B in Figure 11,

then the tote must move the distance on level D and the entire distance on level C, shown as a

at point B will travel longer to get to packing than point A.

Figure 11: Illustration of Conveyor Movement

The location of the last product on the pick list is used to determine the travelling

level A the y co-ordination is 9 blocks and on levels B, C, D there are

Figure 12, the distance from the la

shown as a green arrow, and the distance of which the conveyor is

red arrow, is added together (see

used.

32

: Illustration of Conveyor Movement on Picking Levels

The location of the last product on the pick list is used to determine the travelling

ordination is 9 blocks and on levels B, C, D there are 12 blocks as illustrated in

he distance from the last pick point to the end of the level is calculated

and the distance of which the conveyor is longer, which

(see Figure 12).In essence the y co-ordinate of the final location is

The location of the last product on the pick list is used to determine the travelling distance. If on

12 blocks as illustrated in

st pick point to the end of the level is calculated, which is

longer, which is shown as a

ordinate of the final location is

Figure 12: Level Layout

This calculation is done in Microsoft Excel® wi

picker is situated on. If the picker is on level A then

level A. If the picker is on levels B,

those levels contain 12 blocks. E.g.

the distance marked by the letter A

conveyor is longer, marked with the red arrow

last pick point to the end of the conveyor.

The distances used in the “if” statements are shown in the

33

This calculation is done in Microsoft Excel® with 21 “if” statements. It determines

. If the picker is on level A then a certain distance is used for the 9 blocks on

level A. If the picker is on levels B, C, D then a different distance, than level A is used, because

12 blocks. E.g. If the picker is on level C (see Figure 12) in block 3

the distance marked by the letter A is calculate and then added with the distance which th

, marked with the red arrow. The total y co-ordination is calculated from the

last pick point to the end of the conveyor.

The distances used in the “if” statements are shown in the Table 1.

determines the level the

used for the 9 blocks on

n level A is used, because

) in block 3-5 then

is calculate and then added with the distance which the

ordination is calculated from the

Table 1: Distance Conveyor travel from each block

Block Number in

y co-ordination

1

2

3

4

5

6

7

8

9

10

11

12

After conducting a few measurements on the conveyors system it was concluded that all

conveyor travel at a constant speed.

obtain the time the conveyor takes to move from a level in picking to packing.

the different level is shown in a matrix

Appendix 10.1, Table 20.

4.1.6 Process Time Picking

The process time is the time a picker takes to pick a number of units of a specific business unit

(footwear, apparel or equipment). Intensive time studies were done and

each station. The time to pick footwear differs to the other business units

34

: Distance Conveyor travel from each block

Distance conveyor travel (m)

Block Number in

Level A Level B Level C

6 70 5

12 66 9

18 61 14

24 56 18

30 52 23

36 47 28

42 43 32

48 38 37

54 33 41

29 46

24 51

20 55


conveyor travel at a constant speed. The distance is then multiplied with the speed

obtain the time the conveyor takes to move from a level in picking to packing.

the different level is shown in a matrix form and these distances are read into the

Process Time Picking


(footwear, apparel or equipment). Intensive time studies were done and data was obtained from

each station. The time to pick footwear differs to the other business units due to

Level C


speed in order to

obtain the time the conveyor takes to move from a level in picking to packing. The distance of

read into the simulation see


data was obtained from

due to the irregular

size difference in apparel and equipment

time to pick thus they are combined into

from the time studies were analysed

Table 2: Processing Times for Picking

Process Time (in sec)

Other

Footwear

Using these results mathematical expression

(Appendix 10.2.1 and 10.2.2) which indicates

and maximum as stated in Table

change in the simulation model, thus they are user defined.

4.1.7 Process Time Packing

The process packing time includes all the different business units, no distinction can be made

between footwear and other. The data obtained for packing follows a

developed in Arena® Input Analyzer (

average and the standard deviation

will automatically change in the simulation model, thus they are user defined.

Table 3: Processing Times for Packing

Packing Time

35

in apparel and equipment. Apparel and equipment takes the same amount of

time to pick thus they are combined into a business unit called “Other”. The results obtained

were analysed.

: Processing Times for Picking in sec/unit

Process Time (in sec) Minimum Average Maximum

1 19

3 11

results mathematical expressions were developed in Arena® Input Analyzer

which indicates triangular distributions with the minimum, average

Table 2. These values may be changed and will automatically

change in the simulation model, thus they are user defined.

Process Time Packing


between footwear and other. The data obtained for packing follows a mathematical expression

developed in Arena® Input Analyzer (Appendix 10.2.3) which is a normal distribution

average and the standard deviation are indicated in Table 3. These values may be changed and

will automatically change in the simulation model, thus they are user defined.

: Processing Times for Packing

Average (in sec) Std Dev

287 98.2

. Apparel and equipment takes the same amount of

The results obtained

Maximum

48

26

developed in Arena® Input Analyzer

with the minimum, average

These values may be changed and will automatically


mathematical expression

normal distribution. The

These values may be changed and

4.2 Soft Data

The soft data include all the data relating to the workforce and personnel. The company policies

govern the staffing regulations, such as the number of staff members with their respective skill

levels, daily shift hours and weekly schedules.

Figure 13 is adapted in combining information from Buzacott & Shanthikumar (1993), Kelton,

Sadowski & Sturrock (2007) and Chase, Jacobs and Aquilano (2006) to illu

involved in the simulation study.

Figure 13: Data input to Simulation Study

36



levels, daily shift hours and weekly schedules.

is adapted in combining information from Buzacott & Shanthikumar (1993), Kelton,

Sadowski & Sturrock (2007) and Chase, Jacobs and Aquilano (2006) to illu

: Data input to Simulation Study



is adapted in combining information from Buzacott & Shanthikumar (1993), Kelton,

Sadowski & Sturrock (2007) and Chase, Jacobs and Aquilano (2006) to illustrate the data

5 Depiction of Simulation Model

The development of the simulation model is seen as the most difficult and critical part because

of the detailed description and conceptual translation of the system. Logic relations combines

with mathematical relations must represent the important attribut

simulation model is divided into four

5.1 Conceptual Translation

The conceptual translation of a model lays an essential foundation in the progress or design

phase and is also referred to as identification or the ‘piec

translating a model is to, formulate the model design and approach before one actually

the model in Arena®. The justification behind this is to ensure that all mechanisms and ‘building

components’ are considered. Some of

structure or constraints, the type of analysis to be performed, the type of animation required and

replication length, just to name a few.

5.1.1 Entities

Entities are the dynamic objects of the simulatio

are affected by other entities and the state of the system and affect the output performance

measures (Kelton, Sadowski & Sturrock 2007:20).

A single entity is created at time zero merely to assign a set of at

model. A Read/Write module is used to obtain all

Excel® spreadsheets.

The second entity which is created receives the attributes assigned to it.

replicated and each replication is

5.1.2 Attributes

An attribute is a common characteristic of all entities, but with a specific value that can differ

from one entity to another (Kelton, Sadowski & Sturrock 2007:2

37

Depiction of Simulation Model



with mathematical relations must represent the important attributes of the processes.

tion model is divided into four activity areas.

Conceptual Translation


phase and is also referred to as identification or the ‘pieces of the model’. Conceptual

formulate the model design and approach before one actually


components’ are considered. Some of the things that one should take into account are the data


replication length, just to name a few.

Entities are the dynamic objects of the simulation. They move around, change status, affect and

e affected by other entities and the state of the system and affect the output performance

measures (Kelton, Sadowski & Sturrock 2007:20).

A single entity is created at time zero merely to assign a set of attributes and variables to the

model. A Read/Write module is used to obtain all the attributes and variables from

The second entity which is created receives the attributes assigned to it.

replication is specified as an order made by a customers.


from one entity to another (Kelton, Sadowski & Sturrock 2007:21).



es of the processes. The


es of the model’. Conceptual

formulate the model design and approach before one actually builds


the things that one should take into account are the data


n. They move around, change status, affect and

e affected by other entities and the state of the system and affect the output performance

tributes and variables to the

the attributes and variables from the Microsoft

The second entity which is created receives the attributes assigned to it. This entity gets


By attaching multiple attribute values to each en

of each activity will be listed independently, since their values are determined by an Excel®

spreadsheet. All the attributes needed

Table 4.

Table 4: List of Attributes

Attribute Name

TransTime_A

TransTime_B

TransTime_C

TransTime_D

the picker to walk from one pick

station to the next on a specific

Order#

#beBatched must be batched, if the order is

Qty_LA_Other

Qty_LA_FW

Qty_LB_Other

Qty_LB_FW

Qty_LC_Other

Qty_LC_FW

Qty_LD_Other

Qty_LD_FW

ConvDist_LA

conveyor to the end of the level

ConvDist_LB

ConvDist_LC

ConvDist_LD

38

attribute values to each entity, each entity becomes unique.

activity will be listed independently, since their values are determined by an Excel®

needed to perform the outbound process flow model are listed

Description Purpose

Time (Transport time) it takes

the picker to walk from one pick

station to the next on a specific

level.

To measure the time it takes

to walk, a specific distance by

that operator for an order.

Each order has a specific

number which is call the order

number.

In order to rectify what

customer placed what order.

It indicated how many totes

must be batched, if the order is

split in the picking phase.

To make sure that all the totes

that have been split is

together for the

The number of Other/FW units

that needs to be picked at a

specific level (Level A, B, C or

D)

It is used to be multiplied by

the time to pick that specific

unit. In order to get the total

processes picking time.

The distance the conveyor

needs to travel from where the

picker places the units on the

conveyor to the end of the level

see Figure 12.

To calculate the total time the

conveyor takes to transport

the units to the next station.

The distance is multiplied by

the speed of the conv

each entity becomes unique. The attributes

activity will be listed independently, since their values are determined by an Excel®

model are listed in

Purpose

To measure the time it takes

to walk, a specific distance by

that operator for an order.

In order to rectify what

customer placed what order.

To make sure that all the totes

that have been split is

together for the same order.

It is used to be multiplied by

the time to pick that specific

unit. In order to get the total

processes picking time.

To calculate the total time the

yor takes to transport

to the next station.

ce is multiplied by

of the conveyor.

5.1.3 Variables

A variable is a piece of information that reflects some characteristic of your system, regardless

of how or what kind of entities might be around.

model (Kelton, Sadowski & Sturrock 2007:21).

Table 5: List of Variables

Variable Name

PT_Other_Min

PT_Other_Avg

PT_Other_Max

PT_FW_Min

PT_FW_Avg

PT_FW_Max

PT_Pack_Avg

PT_Pack_StdDev

Input_Taller

Input_Taller + 1

Max#Order

39


of how or what kind of entities might be around. There can be many different variables in a

model (Kelton, Sadowski & Sturrock 2007:21).

Description Purpose

The range of the process

time (picking time). The

process time, is the time a

picker uses to pick a single

unit of type Other.

It is used to calculate the time

a picker takes to pick a

specified number of unit and

of type Other. The quantity is

multiplied by process time.

The range of the process

time (picking time). The

process time, is the time a

picker uses to pick a single

unit of type FW (Footwear).




of type FW. The quantity is


It is the average packing

process time and its standard

deviation.

The packing follows a normal

distribution. The average and

the standard deviation are

used as inputs for the

distribution.

It is a variable that is given

when an entity pass.

It is used to form the logic

that a specified number of

orders are released from the

pick pool for picking.

It is the total number of

orders placed by the pick

pool (the number of the last

order).

It is used to find the last entity

(order) in the simulation

model.


There can be many different variables in a

Purpose




of type Other. The quantity is





of type FW. The quantity is


The packing follows a normal

distribution. The average and

the standard deviation are

used as inputs for the

distribution.

It is used to form the logic

that a specified number of

orders are released from the

pick pool for picking.

is used to find the last entity

(order) in the simulation

model.

The variables used are user-defined and the values are determined by Excel® spreadsheets.

The variables used to perform all the actions in the model are listed in

5.1.4 Resources

Resources represent things like personnel or equipment. An entity seizes (

a resource when available and release

2007:22).

The following resources were identified in the model operations:

Table 6: List of Resources

Resource Name

Picker

Packer A-J

The current system contains 8 picker resource units for the picking and

station, there are 10 pack stations.

40

defined and the values are determined by Excel® spreadsheets.

The variables used to perform all the actions in the model are listed in Table 5.

Resources represent things like personnel or equipment. An entity seizes (a number of

a resource when available and releases it (or them) when finished (Kelton, Sadowski & Sturrock

The following resources were identified in the model operations:

Resource Name Description

It is a person who receives an order

that needs to be picked by hand. The

picker walks to every station and pick

the specific amount of units and place it

in a tote, when finished the tote is

placed on the conveyor.

J

It is a person who packs all the single

units from the totes into boxes for

shipment.

picker resource units for the picking and 2 packers

pack stations.

defined and the values are determined by Excel® spreadsheets.

a number of units of)

s it (or them) when finished (Kelton, Sadowski & Sturrock

It is a person who receives an order

that needs to be picked by hand. The

picker walks to every station and pick

the specific amount of units and place it

in a tote, when finished the tote is

It is a person who packs all the single

boxes for

packers for every pack

5.1.5 Queues

The purpose of a queue is when an entity can not move on because it needs to seize a unit

resource that is tied up by another entity; it needs to be place to wait. (Kelton, Sadowski &

Sturrock 2007:22).That entity will wait in the queue until a resources is available.

The queues defined in the simulation model are listed

Table 7: List of Queues

Queue Name

Transport between Station_LA.Queue

Transport between Station_LB.Queue

Transport between Station_LC.Queue

Transport between Station_LD.Queue

Batch Orders same OrderNumb.Queue

Packer Station_A.Queue

Packer Station_B.Queue

Packer Station_C.Queue

Packer Station_D.Queue

Packer Station_E.Queue

Packer Station_F.Queue

Packer Station_G.Queue

Packer Station_H.Queue

Packer Station_I.Queue

Packer Station_J.Queue

41

The purpose of a queue is when an entity can not move on because it needs to seize a unit



imulation model are listed in Table 7.

Description

Transport between Station_LA.Queue The waiting of a picker resource at a specific

level (A, B, C, D). In order to seize it for

picking at that level.

Transport between Station_LB.Queue

Transport between Station_LC.Queue

Transport between Station_LD.Queue

Batch Orders same OrderNumb.Queue

If an order is split at picking it needs

batched together to ensure that the full order

(not just a part of the order) is despatched to

the customer. It is the waiting of an entity (part

of an order) for the rest of the order to be

completed.

Packer Station_A.Queue

It is the waiting of an entity at the pack station,

for a specific resource, in order to pack the

complete order.

Packer Station_B.Queue

Packer Station_C.Queue

Packer Station_D.Queue

Packer Station_E.Queue

Packer Station_F.Queue

Packer Station_G.Queue

Station_H.Queue

Packer Station_I.Queue

Packer Station_J.Queue

The purpose of a queue is when an entity can not move on because it needs to seize a unit of a



Description

The waiting of a picker resource at a specific

level (A, B, C, D). In order to seize it for

picking at that level.

If an order is split at picking it needs to be

batched together to ensure that the full order

(not just a part of the order) is despatched to

the customer. It is the waiting of an entity (part

of an order) for the rest of the order to be

entity at the pack station,

for a specific resource, in order to pack the

complete order.

5.1.6 Conveyors

Long distance transportation of product is facilitated through the use of belt conveyors systems.

The following conveyors are activities on their own. They were identified to convey totes from

the picking station to the packing station and then from the packing station to the despatching

area. The conveyors in the picking and packing stations move at a

Table 8: Conveyors

Conveyor Name

Conveyor Picking

Conveyor Packing

5.1.7 Picker Travel Pattern

Each order that a picker receive

unique to that specific order) this is shown in

pattern is calculated so that the picking of the units occur in the least amount of walking

distance. The travelling time is proportional to the

the next.

To obtain the travel time of the picker from each station, the distance must be calculated

stated in section 4.1.2. The calculation is done in

model.

Level A is divided into 27 blocks

from the centre of the one block to the next is known and this includes all the blocks on all the

levels. This gives the distance travel by the picker from the one block to the next.

42




. The conveyors in the picking and packing stations move at a constant speed

Conveyor Name Description

Conveyor Picking

It is the time a conveyor uses to

transport totes from the picking

stations to the packing stations.

Conveyor Packing


convey the boxes from the packing

stations to the despatching station.

a picker receive has a unique travel pattern on that level (the travel pattern is

this is shown in Figure 14 and described in section 4.1.2

so that the picking of the units occur in the least amount of walking

is proportional to the distance travelled from one pick

To obtain the travel time of the picker from each station, the distance must be calculated

. The calculation is done in Microsoft Excel® and read in

blocks and levels B, C and D is divided into 36 blocks.


distance travel by the picker from the one block to the next.




speed.


transport totes from the picking

stations to the packing stations.


convey the boxes from the packing

stations to the despatching station.

(the travel pattern is

and described in section 4.1.2. This

so that the picking of the units occur in the least amount of walking

from one pick location to

To obtain the travel time of the picker from each station, the distance must be calculated as

in to the simulation

and levels B, C and D is divided into 36 blocks. The distance


distance travel by the picker from the one block to the next. The system

knows what type of units is stored in each block and a

sequential distance for each order. This gives the travelling distance of a picker betwee

stations and that travelling distance is used to

Figure 14: Picker Travel Pattern

43

stored in each block and a walking pattern is developed for the

order. This gives the travelling distance of a picker betwee

stations and that travelling distance is used to calculate the travelling time of the picker.

walking pattern is developed for the

order. This gives the travelling distance of a picker between

the travelling time of the picker.

6 Simulation Model Construction

The simulation model construction can be very time consuming, depending on its detail and

complexity. The best approach is to start with a simple model and increasing the detail and

complexity as system knowledge

methods used are discussed in this section.

6.1 Model Activity Areas

The simulation model construction is divided into several activity areas to better understand and

illustrate the logic of the model. The activity areas are as follows:

1. Order information generation

2. Pick face

3. Pack stations

4. Despatch and Output transfer

6.1.1 Order Information Generation

In this activity area two entities

variables and the second entity is

The first entity is created at time zero and runs through the read modules

the different variables as in the picking processing time of the differe

processing time and the total number of order

that are imported from the Microsoft Excel® spreadsheets. The variables are then incorporated

into the different distributions for the picking and packing processing times, these distributions

are described in section 4.1.6 and

used at the end of the simulation model, so that the last order writes all the required data to the

Microsoft Excel® spreadsheet.

44

Simulation Model Construction



complexity as system knowledge advance. Verification and validation of the

methods used are discussed in this section.

Model Activity Areas


illustrate the logic of the model. The activity areas are as follows:

formation generation

and Output transfer

Order Information Generation

entities are generated, the first is created to read all the specified

ables and the second entity is specified as an order from a customer.

The first entity is created at time zero and runs through the read modules (Figure

variables as in the picking processing time of the different business units,

processing time and the total number of orders. Section 5.1.3 describes the dif

from the Microsoft Excel® spreadsheets. The variables are then incorporated


and 4.1.7. The variable that indicates the total number of orders is

the simulation model, so that the last order writes all the required data to the



. Verification and validation of the processes and


are generated, the first is created to read all the specified

Figure 15); it reads

nt business units, packing

describes the different variables

from the Microsoft Excel® spreadsheets. The variables are then incorporated


The variable that indicates the total number of orders is

the simulation model, so that the last order writes all the required data to the

Figure 15: Read Simulation Model Variables

The second entity is described as an

enters the first assign module (Figure

The entity then enters a separate module

module where the variable (Input_Taller) is added with a value of one (Input_Taller + 1).

duplicated entity reads (using read module)

value is either a 0 or a 1. If it is a 0 then it means that there are more orders

send from the pick pool (it is not he last order) and if it is a 1 then it is the last order.

decide module verify the value of the number

entity on (the statement is true);

pool. If it is not the last entity then it enters the second decide module.

the pick pool to picking in batches that are specified by the user (e.g.

time to picking). The second decide model

equal to that value which is specified by the user

equal then the entity follows the path explained above

(Input_Taller) is increased with 1.

If the second decide module is true (

module. The scan module scans the queue length at picking (Transport between stations

queues). It verifies if the value of the

to 2, if not then the scan module waits until the queue is smaller or equal to 2. If that occurs then

the entity is released by the scan module. The entity then enters the first assign module, where

the variable (Input_Taller) gets assigned the value zero. The same pr

until the last order is send on by the first decide module.

presentation.

45

: Read Simulation Model Variables

The second entity is described as an order made by a customer and is created at ti

Figure 16) where “zero” is assigned to a variable (Input_Taller).

separate module and is duplicated. It then enters the second assign

module where the variable (Input_Taller) is added with a value of one (Input_Taller + 1).

read module) from a Microsoft Excel® spreadsheet a value,

is either a 0 or a 1. If it is a 0 then it means that there are more orders

send from the pick pool (it is not he last order) and if it is a 1 then it is the last order.

the value of the number (0 or 1) and if that number is 1 then it sends the

; no more entities are duplicated as it is the last order from pick

If it is not the last entity then it enters the second decide module. The order

in batches that are specified by the user (e.g. 100 orders are sent

decide model verify the variable (Input_Taller),

specified by the user, the batch size value (e.g.

ollows the path explained above and the value of

increased with 1. (Input_Taller gets bigger until it equals the batch size value)

If the second decide module is true (equal to the user specified value) then it enters a scan


value of the sum of the queue length divided by four is smaller or equal

then the scan module waits until the queue is smaller or equal to 2. If that occurs then


the variable (Input_Taller) gets assigned the value zero. The same process occur as explained

until the last order is send on by the first decide module. See Figure 16

and is created at time zero. It

a variable (Input_Taller).

enters the second assign

module where the variable (Input_Taller) is added with a value of one (Input_Taller + 1). The

from a Microsoft Excel® spreadsheet a value, this

is either a 0 or a 1. If it is a 0 then it means that there are more orders that needs to be

send from the pick pool (it is not he last order) and if it is a 1 then it is the last order. The first

and if that number is 1 then it sends the

as it is the last order from pick

The orders are sent from

100 orders are sent at a

if the variable is

(e.g. 100). If it is not

and the value of the variable

(Input_Taller gets bigger until it equals the batch size value).

equal to the user specified value) then it enters a scan


divided by four is smaller or equal

then the scan module waits until the queue is smaller or equal to 2. If that occurs then


ocess occur as explained

16 for a graphical

Create single

entity

Figure 16: Duplication of Entity to the Number of Orders

The entity then enters the process of

Excel® spreadsheets. These attributes are order number, quantity and type of business

etc. description of all the input data is described in section

processing sequence, time and

read model. Figure 17 illustrates the path of the entity in obtaining the different attributes of

order. As can be seen in Figure

module is used to assign the current

analyse the simulation time. It gives the entity a time stamp of when it entered the simulation

model.

46

Last Order

: Duplication of Entity to the Number of Orders

The entity then enters the process of getting assigning different attributes from the Microsoft

Excel® spreadsheets. These attributes are order number, quantity and type of business

etc. description of all the input data is described in section 4.1. These attributes will dic

quantity. The attributes are read from the spreadsheet with a

illustrates the path of the entity in obtaining the different attributes of

Figure 17, when the entity enters the simulation model

module is used to assign the current simulation time to that entity, this attribute is later used to


Route to read

in attributes

from the Microsoft

Excel® spreadsheets. These attributes are order number, quantity and type of business units’

. These attributes will dictate

preadsheet with a

illustrates the path of the entity in obtaining the different attributes of each

, when the entity enters the simulation model an assign

attribute is later used to


Route from

Entity creation

Figure 17: Assigning and Reading Attributes to Entity

Appendix 10.4.1 illustrates the complete Arena® model for the order information generation

activity.

6.1.2 Pick Face

The entity assigned with different attributes also known as an order

all the different floors of the pick

and the order is send to all four. A decide module checks the attributed value of the transport

time on that level, if the number is greater

business units occur on that level, thus the order is send to that level

time for that level is zero then it means that no transporting of the business unit

level and the order is send to the dispose module.

47

: Assigning and Reading Attributes to Entity

illustrates the complete Arena® model for the order information generation

The entity assigned with different attributes also known as an order, gets duplicated

all the different floors of the pick face area. There are four levels of picking as stated previously


time on that level, if the number is greater than zero then it means that

on that level, thus the order is send to that level for picking

time for that level is zero then it means that no transporting of the business unit

nd the order is send to the dispose module.

Route to Pick

phase

illustrates the complete Arena® model for the order information generation

gets duplicated and sent to

area. There are four levels of picking as stated previously


than zero then it means that transportation of

for picking. If the transport

time for that level is zero then it means that no transporting of the business unit occurs on that

Route from

Reading

Attributes

Figure 18: Assigning Orders to different levels for picking

After defining that picking should take place on a level by the decide module, the order is send

on for processing of the transport

business units.

The order passes an assign node

as an attribute, this time stamp indicates the time the picking process for an order starts. At the

end of the picking process another assign module assigns the simulation time to that same

order, thus the beginning and ending times of picking is given to the order

write module writes the specified attributes to a Microsoft Excel® spread

analysis of the times an order spends in picking.

48

Route to Pick

Units

Route to Levels

B, C and D

to different levels for picking


on for processing of the transport of the picker from each location and the actual picking of the

The order passes an assign node (Figure 19) where the simulation time is stamped on the order


d of the picking process another assign module assigns the simulation time to that same

order, thus the beginning and ending times of picking is given to the order as

write module writes the specified attributes to a Microsoft Excel® spreadsheet for further

analysis of the times an order spends in picking.

Route to Pick

Units


and the actual picking of the

where the simulation time is stamped on the order


d of the picking process another assign module assigns the simulation time to that same

as attributes. The

sheet for further

Figure 19: Picking and Conveying Processes

The order enters a process model called transport between stations LA (or LB, LC, LD) where a

resource namely a picker is seized and delayed for a specific

the time the picker takes to walk from each location to the next to p

specific order. The time is calculated in Microsoft Excel® see section

calculation of the transport time of a picker for an order.

After the transportation process, the order then gets delayed from the same resource (picker).

The second delay is the process time for the picking of the business units

pick one unit). The quantity of business units is multiplied

specific business unit. The processing time (picking time) for footwear differs from that of the

other business unit, they are added together and form the picking processing time

processing times is described in secti

picked units are placed in totes (

placed on the conveyor and the picker resource is released in order to seize another order.

The conveying of the totes from picking to packing is at a constant velocity, a delay module is

used (Figure 19). The time it takes to convey the totes is directly dependant on the location

where the tote is placed by the picker on the conveyor. The calculation of the distance is

described in section 4.1.5 and the distance the conveyor must travel for that order

49

: Picking and Conveying Processes

a process model called transport between stations LA (or LB, LC, LD) where a

resource namely a picker is seized and delayed for a specific period of time. The delay time is

the time the picker takes to walk from each location to the next to pick business

. The time is calculated in Microsoft Excel® see section 4.1.2

calculation of the transport time of a picker for an order.

, the order then gets delayed from the same resource (picker).

The second delay is the process time for the picking of the business units (actual time it takes to

. The quantity of business units is multiplied with the processing time of that


, they are added together and form the picking processing time

processing times is described in section 4.1.3 refer to that section for more information.

picked units are placed in totes (container for transporting the ordered units), these totes are

the conveyor and the picker resource is released in order to seize another order.


. The time it takes to convey the totes is directly dependant on the location

placed by the picker on the conveyor. The calculation of the distance is

the distance the conveyor must travel for that order

a process model called transport between stations LA (or LB, LC, LD) where a

time. The delay time is

business units for a

4.1.2 for details in

, the order then gets delayed from the same resource (picker).

(actual time it takes to

with the processing time of that


, they are added together and form the picking processing time. The

refer to that section for more information. The

), these totes are

the conveyor and the picker resource is released in order to seize another order.


. The time it takes to convey the totes is directly dependant on the location

placed by the picker on the conveyor. The calculation of the distance is

the distance the conveyor must travel for that order is read into

the model through a read module (

multiplied with the constant velocity of the conveyor which is

time of the conveyor. Appendix

activity.

6.1.3 Pack Station

The tote that comes from the picking area via the conveyor passes an assign node where th

simulation time is given as an attribute to the tote. It is a time stamp of when the packing

process begins. The tote then enter a decide module, the module

specific packing station. The logic is that there are two packer

enters the packing area it must be assign to a pack

busy (if the number of resources busy are less then 2) then the tote is assign to that pack

station otherwise it checks the number of resources

that is less than two then it gets assign to pack station B, this is carried on until pack station J.

Pack station A will be the first to fill then B and so on.

Figure 20: Decide on packing station

50

the model through a read module (Figure 17). The distance which the conveyor

multiplied with the constant velocity of the conveyor which is 4.2sec/meter, giving the total delay

Appendix 10.4.2 illustrates the complete Arena® model for the pick

The tote that comes from the picking area via the conveyor passes an assign node where th


process begins. The tote then enter a decide module, the module must assign the tote to a

specific packing station. The logic is that there are two packers at each packing station, if a tote

a it must be assign to a pack station. First the logic checks if packer A is


he number of resources (packers) currently busy at pack station B


Pack station A will be the first to fill then B and so on.

: Decide on packing station

conveyor must travel is

4.2sec/meter, giving the total delay

illustrates the complete Arena® model for the pick face

The tote that comes from the picking area via the conveyor passes an assign node where the


assign the tote to a

at each packing station, if a tote

station. First the logic checks if packer A is


at pack station B, if


Figure 21 illustrates the actual packing of the units from the totes into boxes for shipment. The

tote enters the packing process and seizes a resources call a packer. The resource is seized,

delayed and released for a certain period of time. This pack processing time is calculated using

a mathematical formula see section

times and the distributions.

Figure 21: Packing Process

The boxes that have been filled with the units are conveyed to

conveyor uses to convey the boxes depend on the

packed. The further away the pack station is from the weigh station the longer the time it takes

the conveyor to convey the boxes.

the conveyor, because it moves at a constant

from the one station to the next.

the end of the packing process is stamped on the box as an attribute.

Appendix 10.4.3 illustrates the complete Arena® model for the pack

51

illustrates the actual packing of the units from the totes into boxes for shipment. The


nd released for a certain period of time. This pack processing time is calculated using

a mathematical formula see section 4.1.6 and 4.1.7 on the analysis of the pack processing

The boxes that have been filled with the units are conveyed to the weigh station. The time the

conveyor uses to convey the boxes depend on the location of pack station where the box was

The further away the pack station is from the weigh station the longer the time it takes

the conveyor to convey the boxes. This distance is calculated and multiplied with the

moves at a constant speed. The delay module is used as the conveyor

The boxes enter an assign module where the simulation time of

of the packing process is stamped on the box as an attribute.

illustrates the complete Arena® model for the pack station activity.

illustrates the actual packing of the units from the totes into boxes for shipment. The


nd released for a certain period of time. This pack processing time is calculated using

he pack processing

the weigh station. The time the

pack station where the box was

The further away the pack station is from the weigh station the longer the time it takes

his distance is calculated and multiplied with the speed of

The delay module is used as the conveyor

The boxes enter an assign module where the simulation time of

activity.


The distance from the weigh stations to the despatching area is a fi

conveying times are constant and the time to weigh a box

time impediment a delay module is used

After the delay the boxes are batched together according to their order number, see section

4.1.4 on how many boxes must be batched per order.

assigned to the boxes; it places a time stamp on how long it took for an order to complete all the

processes. The different attributes are then

spreadsheet. These attributes are the time stamps that were placed on the orders to

time it takes to complete a specified process.

Figure 22: Delay, Batching and Output transfer processes

The box (order) enter a counter and

counter value is equal to the total number of order placed by the pick pool (

Max#Order). If it is true then it is the last order and all the resource utilisations are

through a write module to the Microsoft Excel® spreadsheet. The order is then despatched.

52

Output transfer

stations to the despatching area is a fixed length, thus the

conveying times are constant and the time to weigh a box is constant. Because of the constant

delay module is used to delay the boxes.


on how many boxes must be batched per order. Next an attribute of the simulation time is


processes. The different attributes are then exported with a write module to a Microsoft Excel®

re the time stamps that were placed on the orders to

specified process.

: Delay, Batching and Output transfer processes

The box (order) enter a counter and after that a decide module. The decide module verify if the

counter value is equal to the total number of order placed by the pick pool (

Max#Order). If it is true then it is the last order and all the resource utilisations are

Microsoft Excel® spreadsheet. The order is then despatched.

xed length, thus the

Because of the constant


ute of the simulation time is


with a write module to a Microsoft Excel®

re the time stamps that were placed on the orders to derive the

. The decide module verify if the

counter value is equal to the total number of order placed by the pick pool (the variable:

Max#Order). If it is true then it is the last order and all the resource utilisations are exported

Microsoft Excel® spreadsheet. The order is then despatched.

Route from

Conveyor

Figure 23: Despatching of orders

Appendix 10.4.4 illustrates the complete Arena® model for the despatch and output transfer

activity.

6.2 Verification and Validation of the Model

6.2.1 Verification

Verification is the process of ensuring that the mo

according to the modelling assumptions made.

is an important aspect in verifying the minor details of the simulation model. Debugging is a

major process that is involved in

that the model runs for the desired

6.2.1.1 Verification Techniques

A single entity is sent to enter the model, that entity is followed (step

to be sure that the model logic data is correct.

Animation is used as a visual representation of the activities performed by the model. The use

of various colours, graphs and assigning different picture to the status of each of the reso

(e.g. idle or busy), ensures that the entity flow can be followed.

53

orders

illustrates the complete Arena® model for the despatch and output transfer

Verification and Validation of the Model

Verification is the process of ensuring that the model behaves in the way it was intended,

according to the modelling assumptions made. A thorough investigation on the input parameters


in verification; all problems or bugs need to be cleared to ensure

that the model runs for the desired duration.

Verification Techniques

is sent to enter the model, that entity is followed (step-by-step) through the model

the model logic data is correct.

d as a visual representation of the activities performed by the model. The use

of various colours, graphs and assigning different picture to the status of each of the reso

the entity flow can be followed.

illustrates the complete Arena® model for the despatch and output transfer

del behaves in the way it was intended,

the input parameters


all problems or bugs need to be cleared to ensure

step) through the model

d as a visual representation of the activities performed by the model. The use

of various colours, graphs and assigning different picture to the status of each of the resources

Counters and Discrete-Change Variables

place by consulting the Output Analyz

through counters or variables.

SIMAN code verification is a method of validation. The

Processor (MOD) code can be examined for errors.

6.2.2 Validation

Validation is the task of ensuring that the model behaves the same as the

Validation can be described as the process of ensuring that the built simulation model is a good

representation of the physical syst

Validation does not mean that the model is exactly as the physical system. It does mean that

the model has been built sufficiently,

decisions concerning the physical system by testing them on the simulation model.

In order to validate a simulation model, the results from the model must be compared with the

results from the real system. If the results are similar then the validation is complete.

was done by comparing the outputs

54

Change Variables are also measures of entity flow. Verification takes

by consulting the Output Analyzer. These values should resemble the number of entities

is a method of validation. The Experiment Processor (EXP) and Model

Processor (MOD) code can be examined for errors.

Validation is the task of ensuring that the model behaves the same as the

can be described as the process of ensuring that the built simulation model is a good

of the physical system.


the model has been built sufficiently, representing, the physical system in order to make



. If the results are similar then the validation is complete.

was done by comparing the outputs to the real system outputs.

are also measures of entity flow. Verification takes

er. These values should resemble the number of entities

Experiment Processor (EXP) and Model

Validation is the task of ensuring that the model behaves the same as the physical system.

can be described as the process of ensuring that the built simulation model is a good


representing, the physical system in order to make



. If the results are similar then the validation is complete. Validation

6.3 Scenario Development

The main objective is to test the flexibility of the current system and to see if the demand

for the Soccer World Cup season can be met.

place on the system is also analysed.

scenarios to the simulation models itself in terms of logic mod

parameters imported from Microsoft Excel®.

kept roughly the same with minor changes throughout all the scenarios.

development is divided into three different

were developed.

6.3.1 Increased growth in forth

Information gained regarding growth in the

The purpose of the scenarios is to indicate the impact of growth of the quantity of units order on

the outbound process flow. The growth is

6.3.1.1 Scenario 1: A

System performance and input constraint

next three successive years. The base model is established.

6.3.1.2 Scenario 1: B

Increase the quantity of the units ordered by the customer

effective within three years, in 2012.

6.3.1.3 Scenario 1: C

Based on the changes made in scenario 2, increase the quantity of the units ordered by the

customers with another 5% (a

changes will be effective within three years, in 2015.

55

Scenario Development

The main objective is to test the flexibility of the current system and to see if the demand

for the Soccer World Cup season can be met. Future growth in demand and the pressure it will

is also analysed. There are only a few alterations made in the different

scenarios to the simulation models itself in terms of logic modifications, merely to the input

parameters imported from Microsoft Excel®. Model construction and operational activities are

the same with minor changes throughout all the scenarios.

development is divided into three different sections and from these sections different sce

forthcoming years

Information gained regarding growth in the quantity of units ordered was received from Nike.


the outbound process flow. The growth is in 3 years incremental.

constraint remains the same as the current year

The base model is established.

Increase the quantity of the units ordered by the customers with 5%. These changes will be

effective within three years, in 2012.

de in scenario 2, increase the quantity of the units ordered by the

(a total increase of 10% from the current year 2009)

changes will be effective within three years, in 2015.

The main objective is to test the flexibility of the current system and to see if the demand levels

Future growth in demand and the pressure it will

There are only a few alterations made in the different

ifications, merely to the input

Model construction and operational activities are

the same with minor changes throughout all the scenarios. The scenario

different scenarios

received from Nike.


as the current year (2009) for the

with 5%. These changes will be

de in scenario 2, increase the quantity of the units ordered by the

total increase of 10% from the current year 2009). These

6.3.2 Change in the number of resources

The resources (pickers and packers) are

closely correlated with the number of resources available for a specific process.

number of resources will decrease the resource utilisation but will increas

and this is of more importance to Nike.

compared, refer to section 3.2.

6.3.2.1 Scenario 2: A

The picker is the resource that is places the most strain in the model

pickers available. The picker resource capacity is increased by three to the to

picker of eleven.

6.3.2.2 Scenario 2: B

The picker resource capacity is increased further with four making the total capacity of the

picker resource to fifteen.

6.3.2.3 Scenario 2: C

The number of packer is underutilised

increase the resource utilisation. The number of pack stations is decreased from ten packing

stations to seven packing stations. This is an actual decrease of six packers because each pack

station has two packers assigned to it.

6.3.3 Change the number of

As explained in section 1.1.1 the

before picking tasks are assigned to each picker for processing

orders released from the pick pool is equal to 100

flexibility of the system. The flexibility of the system is an i

system will behave if numerous orders come through of high importance (Expedite orders).

These orders will resemble if a team of the Soccer World Cup goes through to the next round,

then an increase in sales are expected and must be delivered

56

Change in the number of resources

es (pickers and packers) are one of the key factors, the bottlenecks and queues are

closely correlated with the number of resources available for a specific process.

will decrease the resource utilisation but will increase the throughput time

of more importance to Nike. The Key Performance Indicators (KPI’s) will be

The picker is the resource that is places the most strain in the model, because of the number of

. The picker resource capacity is increased by three to the to


underutilised thus a decrease in the number of pack



two packers assigned to it.

the number of orders released from the pick pool

the orders placed by customers are consolidated in a

are assigned to each picker for processing. The batch size of the number of

orders released from the pick pool is equal to 100. Decreasing this value will increase the

flexibility of the system. The flexibility of the system is an important factor, thus te



then an increase in sales are expected and must be delivered within the next day.

, the bottlenecks and queues are

closely correlated with the number of resources available for a specific process. Increasing the

e the throughput time

The Key Performance Indicators (KPI’s) will be

, because of the number of

. The picker resource capacity is increased by three to the total number of


pack stations will



consolidated in a pick pool

The batch size of the number of

. Decreasing this value will increase the

mportant factor, thus testing how the



within the next day. The batch

size is changed from 100 order to 50 order

expedite orders can be handled more efficiently.

6.4 Possible follow-up S

6.4.1 Business unit locations

The distance a picker must walk in the picking station is relative to the

that must be picked. Fast moving items are a

conveyor on the ground floor, level A

level is a way to reduce travel time.

patterns will have a positive effect on the throughput tim

6.4.2 Conveyor Speed

The conveyor takes a very long time in conveying the units

pack station and then to dispatching. Thus analysis of the time the conveyor takes can be

measured and a trade-off study developed. The trade

throughput time of faster conveyors to the capital cost of new conveyors.

57

is changed from 100 order to 50 orders, thus increasing the flexibility of the system and

expedite orders can be handled more efficiently.

up Scenarios

Business unit locations

The distance a picker must walk in the picking station is relative to the classification

Fast moving items are assigned to locations as close as possible to the

conveyor on the ground floor, level A. Different patterns in the way the units are stored on each

level is a way to reduce travel time. Calculating the different travel times for the different

patterns will have a positive effect on the throughput time. Thus, optimising pick directives.

takes a very long time in conveying the units from pick stations to their assigned

station and then to dispatching. Thus analysis of the time the conveyor takes can be

off study developed. The trade-off study is between the incre

throughput time of faster conveyors to the capital cost of new conveyors.

, thus increasing the flexibility of the system and

classification of the items

s possible to the

the way the units are stored on each

Calculating the different travel times for the different

Thus, optimising pick directives.

ations to their assigned

station and then to dispatching. Thus analysis of the time the conveyor takes can be

off study is between the increases in the

7 Simulation Results

All the scenario runs consisted of one replication for the duration of completion of all the orders.

For information regarding the input data refer to section

section 6.3 and will be evaluated in the following section

from reports created by Arena® and

containing a statistical summary of specified data collection are generated after each simulation

run. Additional output date is shown in section Appendix

7.1 Increase growth in forthcoming years

The aim is to evaluate yearly throughput time

order to make correct assumptions.

7.1.1 Scenario 1: A

All constraints remain the same, thus the current system results is obtained in

the current output of Nike. It is also referred to t

Table 9: Base model results

Statistics

Throughput

Time

Time in

Process

58

Simulation Results and Recommendations


the input data refer to section 4. The scenarios are described in

and will be evaluated in the following section regarding output parameters attained

from reports created by Arena® and data exported to Microsoft Excel® spreadsheets.


dditional output date is shown in section Appendix 10.3.

Increase growth in forthcoming years

The aim is to evaluate yearly throughput time of the orders and utilisation in each section in

order to make correct assumptions.

All constraints remain the same, thus the current system results is obtained in

lso referred to the base model.

Statistics Output Value (min)

Throughput Average 89.55

Average 79.03

Picking LA 90.65

Picking LB 77.62

Picking LC 54.34

Picking LD 60.3

Packing 6.16


rios are described in

output parameters attained

spreadsheets. Reports


and utilisation in each section in

All constraints remain the same, thus the current system results is obtained in Table 9. It states

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Packer

A

Packer

B

Packer

C

58.3%

48.9%

37.6%

Av

era

ge

Resource Utilisation

Figure 24: Resource utilisation of base model

The picker is the resource that is utilised the most

(bottle neck).

7.1.2 Scenario 1: B

As expected the increase in the

throughput time and also the time in process.

59

Packer Packer

D

Packer

E

Packer

F

Packer

G

Packer

H

Packer

I

Packer

J

Picker

37.6%

23.8%

13.8%

7.1%2.2% 0.8% 0.2% 0.0%

77.0%


: Resource utilisation of base model with 8 pickers

The picker is the resource that is utilised the most, indicating that it is the constraint resource

As expected the increase in the number of units ordered by the customers increased the

throughput time and also the time in process.

Picker

77.0%

, indicating that it is the constraint resource

number of units ordered by the customers increased the

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Packer

A

Packer

B

Packer

C

60.2%

48.5%

34.6%

Av

era

ge


Table 10: Three year projection in quantity

Statistics

Throughput

Time

Time in

Process

The utilisation of the resources increased, with

5% increase of the number of units ordered by the customers.

Figure 25: Resource utilisation of three

60

Packer

D

Packer

E

Packer

F

Packer

G

Packer

H

Packer I Packer J

21.0%

11.8%

5.8%2.1% 0.8% 0.0% 0.0%


: Three year projection in quantity of units ordered results



Average 84.41

Picking LA 64.16

Picking LB 57.93

Picking LC 82.52

Picking LD 97.07

Packing 6.13

the resources increased, with more work that had to be done because of

units ordered by the customers.

: Resource utilisation of three year projection

Packer J Picker

0.0%

80.0%

re work that had to be done because of the

It is suggested that over the period of three years with an increase of 5% on the number of units

ordered by the customer, that the number of the pickers must be increased.

can be done on the number of picker resources required in order

throughput time. It is not necessary to increase the number of packer because the last four

packing stations utilisation’s is almost zero.

7.1.3 Scenario 1: C

As in the previous scenario the

increase of the number of units ordered from customers.

Table 11: Six year projection in quantity of units ordered

Statistics

Throughput

Time

Time in

Process

An increase in volume results in higher resource utilisation.

61


ordered by the customer, that the number of the pickers must be increased. Additional studies

can be done on the number of picker resources required in order to maintain the current

t is not necessary to increase the number of packer because the last four

packing stations utilisation’s is almost zero.

the previous scenario the throughput time and the time in process incre

increase of the number of units ordered from customers.

: Six year projection in quantity of units ordered results



Average 89.1

Picking LA 102.65

Picking LB 86.13

Picking LC 62.37

Picking LD 67.49

Packing 6.15

An increase in volume results in higher resource utilisation.


Additional studies

to maintain the current

t is not necessary to increase the number of packer because the last four

time and the time in process increases with the

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Packer

A

Packer

B

Packer

C

64.4%

53.0%

36.5%

Av

era

ge


Figure 26: Resource utilisation of six year projection

As the year pass more units are ordered thus more pickers must be employed to have the same

throughput time as the current year.

resources required in order to maintain the current throughput time.

62

Packer Packer

D

Packer

E

Packer

F

Packer

G

Packer

H

Packer

I

Packer

J

Picker

36.5%

21.7%

10.6%4.7%

1.5% 0.5% 0.1% 0.0%

86.2%


: Resource utilisation of six year projection

year pass more units are ordered thus more pickers must be employed to have the same

throughput time as the current year. Additional studies can be done on the number of picker

resources required in order to maintain the current throughput time.

Picker

86.2%

year pass more units are ordered thus more pickers must be employed to have the same

Additional studies can be done on the number of picker

7.2 Change in resource capacity

The aim is to evaluate the throughput time of the orders and utilisation with the change in

resources. Refer to section 7.1.1

7.2.1 Scenario 2: A

The increase in the number of picker resources greatly reduced the throughput time. This is an

indication of the sensitivity of the sy

in all the times is illustrated.

Table 12: Increase picker resource capacity

Statistics

Throughput

Time

Time in

Process

The resource utilisation for the picker shows a

utilisation of the packers (except for packer A). This change in the utilisation is as predicted,

more picker are available thus more units

the packers work harder because

more important for Nike to decrease the throughput time

utilisations. In doing this more units can be sold and in turn more revenue.

63

resource capacity


7.1.1 for the base case of the simulation model.


indication of the sensitivity of the system related to the number of picker resources. A reduction

: Increase picker resource capacity by three results



Average 62.94

Picking LA 70

Picking LB 63.49

Picking LC 44.93

Picking LD 52.31

Packing 6.03

for the picker shows a decrease, but an increase is shown in the


more picker are available thus more units are pushed through the system. This in turn makes

s work harder because the units are sent from picking at a much higher rate.

more important for Nike to decrease the throughput time then to increase all the resource

. In doing this more units can be sold and in turn more revenue.



stem related to the number of picker resources. A reduction

but an increase is shown in the


his in turn makes

the units are sent from picking at a much higher rate. It is

then to increase all the resource

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Packer

A

Packer

B

Packer

C

55.0%

49.2%

41.6%

Av

era

ge


Figure 27: Resource utilisation with increase of three pickers

7.2.2 Scenario 2: B

Increasing the picker also indicated a reduction

The results indicate a drop in picker

increase in the packer utilisation indicates that more units are pushed through the system.

64

Packer Packer

D

Packer

E

Packer

F

Packer

G

Packer

H

Packer

I

Packer

J

Picker

41.6%

32.1%

22.1%

14.6%

8.8%4.7%

2.6% 1.4%

68.3%


: Resource utilisation with increase of three pickers

also indicated a reduction in the throughput time and the time in process.

results indicate a drop in picker utilisation compared to pervious scenarios


Picker

68.3%

in the throughput time and the time in process.

scenarios, but a greater


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Packer

A

Packer

B

Packer

C

68.5%63.8%

57.7%

Av

era

ge

Table 13: Increase picker resource capacity by seven results

Statistics

Throughput

Time

Time in

Process

The results obtained behaved as the previous scenario.

decreased the throughput time.

utilisation except for the packer resource.

Figure 28: Resource utilisation with increase of seven pic

65

Packer Packer

D

Packer

E

Packer

F

Packer

G

Packer

H

Packer

I

Packer

J

57.7%

51.0%

41.3%

31.9%

23.4%

15.3%10.7%

13.5%


: Increase picker resource capacity by seven results



Average 48.88

Picking LA 52.79

Picking LB 50.89

Picking LC 35.68

Picking LD 43.4

Packing 6.15

The results obtained behaved as the previous scenario. Increasing the resource capacity

decreased the throughput time. The values obtained are lower in the times and increase in the

utilisation except for the packer resource.

: Resource utilisation with increase of seven pickers

Packer Picker

13.5%

69.8%

Increasing the resource capacity

and increase in the

It is suggested that Nike employ more picker. More analysis can be done on the

but increasing the number of pickers

an increase in the utilisation of the packers

7.2.3 Scenario 2: C

Decreasing the number of packing station by three did not have an adverse effect on the system

as predicted. This is because the base

stations is minimal.

Table 14: Decrease packer resource capacity by three results

Statistics

Throughput

Time

Time in

Process

66

Nike employ more picker. More analysis can be done on the

the number of pickers shows a considerable decrease in the throughput time and

of the packers.


as predicted. This is because the base case shows that the utilisation for the last packing

packer resource capacity by three results



Average 78.06

Picking LA 89.05

Picking LB 76.59

Picking LC 54.65

Picking LD 60.63

Packing 6.26

Nike employ more picker. More analysis can be done on the exact number,

in the throughput time and


utilisation for the last packing

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Packer A Packer B

57.5%

49.1%

Av

era

ge


Figure 29: Resource utilisation with decrease of three packing stations

This illustrates that the number of packing station

with very minimal effect on the throughput time and the

the packing station only increases their utilisation with a small percentage to absorb the work of

the packing stations that have been taken away

67

Packer C Packer D Packer E Packer F Packer G Picker

38.2%

24.5%

15.1%

7.7%4.9%

76.9%


: Resource utilisation with decrease of three packing stations

This illustrates that the number of packing stations may be reduced with up to as many as three,

with very minimal effect on the throughput time and the utilisation of the resources. The rest of


the packing stations that have been taken away.

Picker

76.9%

ced with up to as many as three,

utilisation of the resources. The rest of


7.3 Change the batch size of

The aim is to evaluate the throughput time and

the number of orders send from the pick

batch sizes reduce the total system work in progress

Decreasing the number of orders from, the pick pool that is

flexibility of the system. Expedite orders are dealt with in a minimal amount of time, increasing

the customer satisfaction and meeting with the high

Figure 30: Decrease in orders send

Statistics

Throughput

Time

Time in

Process

The throughput time and the time in process are reduced with the reduction of the number of

orders from the pick pool and the utilisation increased considerably. The last number of packer

show a zero or very low utilisation, this indicates

68

batch size of orders send from the pick pool

m is to evaluate the throughput time and the utilisation of the orders with the change in

send from the pick pool; this is done to test the systems flexibility

batch sizes reduce the total system work in progress

e number of orders from, the pick pool that is released for picking, increases the


the customer satisfaction and meeting with the high needed flexibility of the Soccer World Cup.

s send from pick pool results



Average 52.76

Picking LA 55.99

Picking LB 55.56

Picking LC 39.88

Picking LD 48.26

Packing 6.12


orders from the pick pool and the utilisation increased considerably. The last number of packer

show a zero or very low utilisation, this indicates that less packing stations are required.

send from the pick pool

with the change in

flexibility. Smaller

picking, increases the


the Soccer World Cup.


orders from the pick pool and the utilisation increased considerably. The last number of packers

less packing stations are required.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Packer

A

Packer

B

Packer

C

67.7%

55.8%

41.4%

Av

era

ge


Figure 31: Resource utilisation with decrease in order

It is suggested that Nike reduce the

will increase the flexibility which

Cup.

As stated previously it is more important for Nike to decrease the throughput time than to

increase the utilisation of the resources. If more units are delivered to the customers more units

will be sold and so increasing the revenue of Nike.

The results obtained through the scenario development indicate that the simulation model is a

true representation of the Nike’s

output results as predicted. The simulation model may be used to test and analyse several

different scenarios.

69

Packer Packer

D

Packer

E

Packer

F

Packer

G

Packer

H

Packer

I

Packer

J

Picker

41.4%

27.7%

16.5%

8.4%3.6% 1.4% 0.0% 0.0%

87.1%


: Resource utilisation with decrease in orders send from pick pool

reduce the batch size of the number of orders released for picking

ill increase the flexibility which is especially important for the up and coming Soccer World


e the utilisation of the resources. If more units are delivered to the customers more units

will be sold and so increasing the revenue of Nike.


’s outbound process flow. The input parameters influenced the


Picker

87.1%

size of the number of orders released for picking. This

is especially important for the up and coming Soccer World


e the utilisation of the resources. If more units are delivered to the customers more units


outbound process flow. The input parameters influenced the


8 Conclusion

This project not only satisfied the objectives of increas

the demand of the Soccer World Cup will be met. The managing and understanding of the

intricate processes of the Nike distribution facility will lead to higher service levels, satisfied

customers and an efficient workforce.

The most suitable industrial engineering tool to assists the management in controlling the

factors that affect Nike’s performance is the use of simulation modelling. The literature study

point out the Arena® simulation model is the most suitabl

The conceptual design combined with the stated input parameter and business decisions

ensure that all of the relevant matters are included in the model.

The unstable demand of the Soccer World Cup period is

influence the structure of the simulation model.

represents the real world and several scenarios are tested.

execution of the project can be use

resources deployment will save Nike labour cost due to high productivity and utilisation in

different department. The system changes will reduce process cycle time and order lead time,

which in turn will result in increased customer satisfaction. Increasing the flexibility indicated that

more goods could be delivered in the same period of time, thus increasing the number of sales.

Through successful identification of the bottlenecks in current

able to recognise and prioritise investment opportunities. By investing in the most appropriate

areas, Nike will not only improve their plant utilisation

chasing the root cause for the plant utilisation optimisation.

Overall the project is a huge success and will yield significant cost reduction after

implementation in February 2010. Preliminary indications are that the project will yield a

reduction on unit processing costs

re-usable internally and further studies can be conducted on outbound process flow and

planning processes.

70

This project not only satisfied the objectives of increasing plant flexibility, but also proved that



workforce.



point out the Arena® simulation model is the most suitable method for completing the project.


ensure that all of the relevant matters are included in the model.

demand of the Soccer World Cup period is some of the considerations that

influence the structure of the simulation model. After verification and validation the model

represents the real world and several scenarios are tested. The scenarios generated during the

execution of the project can be used to address the initial problems identified. The change in the



n turn will result in increased customer satisfaction. Increasing the flexibility indicated that


Through successful identification of the bottlenecks in current operational activities, Nike will be


areas, Nike will not only improve their plant utilisation and flexibility, but will also save money by

lant utilisation optimisation.



reduction on unit processing costs and increase throughput capacity. The simulation model is

usable internally and further studies can be conducted on outbound process flow and

ing plant flexibility, but also proved that





e method for completing the project.


some of the considerations that

erification and validation the model

The scenarios generated during the

The change in the



n turn will result in increased customer satisfaction. Increasing the flexibility indicated that


operational activities, Nike will be


, but will also save money by



The simulation model is

usable internally and further studies can be conducted on outbound process flow and

9 References

ACSL. [online] http://www.acslsim.

Advanced continuous simulation language. [online] sim.sagepub.com.innopac.up.ac.za/cgi/content/abstract/26/3/72

Arena® Professional Edition. [online] http://www.arenasimulation.com/products/professional_edition.asp

Bowersox, DJ & Marien, EJ & Helfericdistribution systems: Discussion of the computer model

http://www.acslsim.com/products/modeling.shtml

Buzacott, JA & Shanthikumar, JG 1993, Hall, New Jersey.

Chase, RB & Jacobs, FR & Aquilano, NJ 2006, advantage with global cases, 11

Harrington, HJ & Tumay, K 2000,

Kelton, WD & Sadowski, RP & Sturrock, DT 2007, New York.

Klingstam, P & Gullander, P 1999, engineering, Elsevier Science B.V., Sweden. [online] http://www.sciencedirect.com.innopac.up.ac.za/science

71

http://www.acslsim.com/products, accessed on 13May 2009

Advanced continuous simulation language. [online] http://o-sim.sagepub.com.innopac.up.ac.za/cgi/content/abstract/26/3/72, accessed on 13May 2009

Professional Edition. [online] http://www.arenasimulation.com/products/professional_edition.asp, accessed on 14May 2009

Bowersox, DJ & Marien, EJ & Helferich, OK, Dynamic simulation model and planning physical

distribution systems: Discussion of the computer model. [online] http://www.acslsim.com/products/modeling.shtml, accessed on 11May 2009

Buzacott, JA & Shanthikumar, JG 1993, Stochastic models of manufacturing systems

Chase, RB & Jacobs, FR & Aquilano, NJ 2006, Operations management for competitive

, 11th edn, McGraw-Hill, New York.

arrington, HJ & Tumay, K 2000, Simulation modeling methods, McGraw-Hill, New York.

Kelton, WD & Sadowski, RP & Sturrock, DT 2007, Simulation with Arena, 4th edn, McGraw

Klingstam, P & Gullander, P 1999, Overview of simulation tools for computer-aided production

, Elsevier Science B.V., Sweden. [online] http://www.sciencedirect.com.innopac.up.ac.za/science, accessed on 13May 2009

13May 2009

accessed on 14May 2009

Dynamic simulation model and planning physical

Stochastic models of manufacturing systems, Prentice-

Operations management for competitive

Hill, New York.

edn, McGraw-Hill,

aided production


Mehrotra, V., Fama, J., 2003.

opportunities. Proceedings of the 2003 Winter Simulation Conference. Edited by: Chick,

S.,Sanchez,P.J.,Ferrin,D.,Morrice, D.J.

ProModel Optimisation Software. [online] http://www.promodel.com/products/promodel/features.asp

Wikipedia – Acsl. [online] http://en.wikipedia.org/wiki/advanced_continuous_simulation_language2009

Wikipedia - Arena Rockwell Software. [online] accessed on 14May 2009

Winston, WL 2004, Introduction to probability models, operations research,

Thomson Brooks/Cole, Canada.

72

Mehrotra, V., Fama, J., 2003. Call centre simulation modelling: methods, challenges and

Proceedings of the 2003 Winter Simulation Conference. Edited by: Chick,

S.,Sanchez,P.J.,Ferrin,D.,Morrice, D.J.

ProModel Optimisation Software. [online] http://www.promodel.com/products/promodel/features.asp, accessed on 13May 2009

wiki/advanced_continuous_simulation_language, accessed on 11May

Arena Rockwell Software. [online] http://en.wikipedia.org/wiki/Rockwell_Arena

Introduction to probability models, operations research, vol.2,

Call centre simulation modelling: methods, challenges and

Proceedings of the 2003 Winter Simulation Conference. Edited by: Chick,


accessed on 11May

http://en.wikipedia.org/wiki/Rockwell_Arena,

vol.2, 4th edn,

Sum

of T

rans

port

Tim

eCo

lum

n La

bels

Row

Lab

els

1-1

1-2

1-3

1-4

1-5

1-6

1-7

1-8

1-9

2-1

2-2

2-3

2-4

2-5

2-6

2-7

2-8

2-9

3-1

3-2

3-3

3-4

3-5

3-6

3-7

3-8

3-9

Gra

nd T

otal

1-1

015

3045

6075

9010

512

011

2641

5671

8610

111

613

122

3752

6782

9711

212

714

219

17

1-2

150

1530

4560

7590

105

2611

2641

5671

8610

111

637

2237

5267

8297

112

127

1602

1-3

3015

015

3045

6075

9041

2611

2641

5671

8610

152

3722

3752

6782

9711

213

77

1-4

4530

150

1530

4560

7556

4126

1126

4156

7186

6752

3722

3752

6782

9712

42

1-5

6045

3015

015

3045

6071

5641

2611

2641

5671

8267

5237

2237

5267

8211

97

1-6

7560

4530

150

1530

4586

7156

4126

1126

4156

9782

6752

3722

3752

6712

42

1-7

9075

6045

3015

015

3010

186

7156

4126

1126

4111

297

8267

5237

2237

5213

77

1-8

105

9075

6045

3015

015

116

101

8671

5641

2611

2612

711

297

8267

5237

2237

1602

10 Appendices

10.1 Input Data Analysis

Table 15: Picker Transport Time Table for Level A

73

1-8

105

9075

6045

3015

015

116

101

8671

5641

2611

2612

711

297

8267

5237

2237

1602

1-9

120

105

9075

6045

3015

013

111

610

186

7156

4126

1114

212

711

297

8267

5237

2219

17

2-1

1126

4156

7186

101

116

131

015

3045

6075

9010

512

011

2641

5671

8610

111

613

118

18

2-2

2611

2641

5671

8610

111

615

015

3045

6075

9010

526

1126

4156

7186

101

116

1503

2-3

4126

1126

4156

7186

101

3015

015

3045

6075

9041

2611

2641

5671

8610

112

78

2-4

5641

2611

2641

5671

8645

3015

015

3045

6075

5641

2611

2641

5671

8611

43

2-5

7156

4126

1126

4156

7160

4530

150

1530

4560

7156

4126

1126

4156

7110

98

2-6

8671

5641

2611

2641

5675

6045

3015

015

3045

8671

5641

2611

2641

5611

43

2-7

101

8671

5641

2611

2641

9075

6045

3015

015

3010

186

7156

4126

1126

4112

78

2-8

116

101

8671

5641

2611

2610

590

7560

4530

150

1511

610

186

7156

4126

1126

1503

2-9

131

116

101

8671

5641

2611

120

105

9075

6045

3015

013

111

610

186

7156

4126

1118

18

3-1

2237

5267

8297

112

127

142

1126

4156

7186

101

116

131

015

3045

6075

9010

512

019

17

3-2

3722

3752

6782

9711

212

726

1126

4156

7186

101

116

150

1530

4560

7590

105

1602

3-3

5237

2237

5267

8297

112

4126

1126

4156

7186

101

3015

015

3045

6075

9013

77

3-4

6752

3722

3752

6782

9756

4126

1126

4156

7186

4530

150

1530

4560

7512

42

3-5

8267

5237

2237

5267

8271

5641

2611

2641

5671

6045

3015

015

3045

6011

97

3-6

9782

6752

3722

3752

6786

7156

4126

1126

4156

7560

4530

150

1530

4512

42

3-7

112

9782

6752

3722

3752

101

8671

5641

2611

2641

9075

6045

3015

015

3013

77

3-8

127

112

9782

6752

3722

3711

610

186

7156

4126

1126

105

9075

6045

3015

015

1602

3-9

142

127

112

9782

6752

3722

131

116

101

8671

5641

2611

120

105

9075

6045

3015

019

17

Gra

nd T

otal

1917

1602

1377

1242

1197

1242

1377

1602

1917

1818

1503

1278

1143

1098

1143

1278

1503

1818

1917

1602

1377

1242

1197

1242

1377

1602

1917

3952

8

Input Data Analysis

: Picker Transport Time Table for Level A

Gra

nd T

otal

1917

1602

1377

1242

1197

1242

1377

1602

1917

1818

1503

1278

1143

1098

1143

1278

1503

1818

1917

1602

1377

1242

1197

1242

1377

1602

1917

3952

8

Su

m o

f T

ran

spo

rtT

ime

Co

lum

n L

ab

els

Ro

w L

ab

els

1-1

1-1

01

-11

1-1

21

-21

-31

-41

-51

-61

-71

-81

-92

-12

-10

2-1

12

-12

2-2

2-3

2-4

2-5

2-6

2-7

2-8

2-9

3-1

3-1

03

-11

3-1

23

-23

-33

-43

-53

-63

-73

-83

-9G

ran

d T

ota

l

1-1

01

17

13

01

43

13

26

39

52

65

78

91

10

41

41

31

14

41

57

27

40

53

66

79

92

10

51

18

28

14

51

58

17

14

15

46

78

09

31

06

11

91

32

30

78

1-1

01

17

01

32

61

04

91

78

65

52

39

26

13

13

11

42

74

01

18

10

59

27

96

65

34

02

71

45

28

41

54

13

21

19

10

69

38

06

75

44

12

37

6

1-1

11

30

13

01

31

17

10

49

17

86

55

23

92

61

44

27

14

27

13

11

18

10

59

27

96

65

34

01

58

41

28

41

14

51

32

11

91

06

93

80

67

54

26

88

1-1

21

43

26

13

01

30

11

71

04

91

78

65

52

39

15

74

02

71

41

44

13

11

18

10

59

27

96

65

31

71

54

41

28

15

81

45

13

21

19

10

69

38

06

73

07

8

1-2

13

10

41

17

13

00

13

26

39

52

65

78

91

27

11

81

31

14

41

42

74

05

36

67

99

21

05

41

13

21

45

15

82

84

15

46

78

09

31

06

11

92

68

8

1-3

26

91

10

41

17

13

01

32

63

95

26

57

84

01

05

11

81

31

27

14

27

40

53

66

79

92

54

11

91

32

14

54

12

84

15

46

78

09

31

06

23

76

1-4

39

78

91

10

42

61

30

13

26

39

52

65

53

92

10

51

18

40

27

14

27

40

53

66

79

67

10

61

19

13

25

44

12

84

15

46

78

09

32

14

2

Table 16: Picker Transport Time Table for Level B, C, D

74

1-4

39

78

91

10

42

61

30

13

26

39

52

65

53

92

10

51

18

40

27

14

27

40

53

66

79

67

10

61

19

13

25

44

12

84

15

46

78

09

32

14

2

1-5

52

65

78

91

39

26

13

01

32

63

95

26

67

99

21

05

53

40

27

14

27

40

53

66

80

93

10

61

19

67

54

41

28

41

54

67

80

19

86

1-6

65

52

65

78

52

39

26

13

01

32

63

97

96

67

99

26

65

34

02

71

42

74

05

39

38

09

31

06

80

67

54

41

28

41

54

67

19

08

1-7

78

39

52

65

65

52

39

26

13

01

32

69

25

36

67

97

96

65

34

02

71

42

74

01

06

67

80

93

93

80

67

54

41

28

41

54

19

08

1-8

91

26

39

52

78

65

52

39

26

13

01

31

05

40

53

66

92

79

66

53

40

27

14

27

11

95

46

78

01

06

93

80

67

54

41

28

41

19

86

1-9

10

41

32

63

99

17

86

55

23

92

61

30

11

82

74

05

31

05

92

79

66

53

40

27

14

13

24

15

46

71

19

10

69

38

06

75

44

12

82

14

2

2-1

14

13

11

44

15

72

74

05

36

67

99

21

05

11

80

11

71

30

14

31

32

63

95

26

57

89

11

04

14

13

11

44

15

72

74

05

36

67

99

21

05

11

82

91

0

2-1

01

31

14

27

40

11

81

05

92

79

66

53

40

27

11

70

13

26

10

49

17

86

55

23

92

61

31

31

14

27

40

11

81

05

92

79

66

53

40

27

22

08

2-1

11

44

27

14

27

13

11

18

10

59

27

96

65

34

01

30

13

01

31

17

10

49

17

86

55

23

92

61

44

27

14

27

13

11

18

10

59

27

96

65

34

02

52

0

2-1

21

57

40

27

14

14

41

31

11

81

05

92

79

66

53

14

32

61

30

13

01

17

10

49

17

86

55

23

91

57

40

27

14

14

41

31

11

81

05

92

79

66

53

29

10

2-2

27

11

81

31

14

41

42

74

05

36

67

99

21

05

13

10

41

17

13

00

13

26

39

52

65

78

91

27

11

81

31

14

41

42

74

05

36

67

99

21

05

25

20

2-3

40

10

51

18

13

12

71

42

74

05

36

67

99

22

69

11

04

11

71

30

13

26

39

52

65

78

40

10

51

18

13

12

71

42

74

05

36

67

99

22

20

8

2-4

53

92

10

51

18

40

27

14

27

40

53

66

79

39

78

91

10

42

61

30

13

26

39

52

65

53

92

10

51

18

40

27

14

27

40

53

66

79

19

74

2-5

66

79

92

10

55

34

02

71

42

74

05

36

65

26

57

89

13

92

61

30

13

26

39

52

66

79

92

10

55

34

02

71

42

74

05

36

61

81

8

2-6

79

66

79

92

66

53

40

27

14

27

40

53

65

52

65

78

52

39

26

13

01

32

63

97

96

67

99

26

65

34

02

71

42

74

05

31

74

0

2-7

92

53

66

79

79

66

53

40

27

14

27

40

78

39

52

65

65

52

39

26

13

01

32

69

25

36

67

97

96

65

34

02

71

42

74

01

74

0

2-8

10

54

05

36

69

27

96

65

34

02

71

42

79

12

63

95

27

86

55

23

92

61

30

13

10

54

05

36

69

27

96

65

34

02

71

42

71

81

8

2-9

11

82

74

05

31

05

92

79

66

53

40

27

14

10

41

32

63

99

17

86

55

23

92

61

30

11

82

74

05

31

05

92

79

66

53

40

27

14

19

74

3-1

28

14

51

58

17

14

15

46

78

09

31

06

11

91

32

14

13

11

44

15

72

74

05

36

67

99

21

05

11

80

11

71

30

14

31

32

63

95

26

57

89

11

04

30

78

3-1

01

45

28

41

54

13

21

19

10

69

38

06

75

44

11

31

14

27

40

11

81

05

92

79

66

53

40

27

11

70

13

26

10

49

17

86

55

23

92

61

32

37

6

3-1

11

58

41

28

41

14

51

32

11

91

06

93

80

67

54

14

42

71

42

71

31

11

81

05

92

79

66

53

40

13

01

30

13

11

71

04

91

78

65

52

39

26

26

88

3-1

21

71

54

41

28

15

81

45

13

21

19

10

69

38

06

71

57

40

27

14

14

41

31

11

81

05

92

79

66

53

14

32

61

30

13

01

17

10

49

17

86

55

23

93

07

8

3-2

41

13

21

45

15

82

84

15

46

78

09

31

06

11

92

71

18

13

11

44

14

27

40

53

66

79

92

10

51

31

04

11

71

30

01

32

63

95

26

57

89

12

68

8

3-3

54

11

91

32

14

54

12

84

15

46

78

09

31

06

40

10

51

18

13

12

71

42

74

05

36

67

99

22

69

11

04

11

71

30

13

26

39

52

65

78

23

76

3-4

67

10

61

19

13

25

44

12

84

15

46

78

09

35

39

21

05

11

84

02

71

42

74

05

36

67

93

97

89

11

04

26

13

01

32

63

95

26

52

14

2

3-5

80

93

10

61

19

67

54

41

28

41

54

67

80

66

79

92

10

55

34

02

71

42

74

05

36

65

26

57

89

13

92

61

30

13

26

39

52

19

86

3-6

93

80

93

10

68

06

75

44

12

84

15

46

77

96

67

99

26

65

34

02

71

42

74

05

36

55

26

57

85

23

92

61

30

13

26

39

19

08

3-7

10

66

78

09

39

38

06

75

44

12

84

15

49

25

36

67

97

96

65

34

02

71

42

74

07

83

95

26

56

55

23

92

61

30

13

26

19

08

3-8

11

95

46

78

01

06

93

80

67

54

41

28

41

10

54

05

36

69

27

96

65

34

02

71

42

79

12

63

95

27

86

55

23

92

61

30

13

19

86

3-9

13

24

15

46

71

19

10

69

38

06

75

44

12

81

18

27

40

53

10

59

27

96

65

34

02

71

41

04

13

26

39

91

78

65

52

39

26

13

02

14

2

Gra

nd

To

tal

30

78

23

76

26

88

30

78

26

88

23

76

21

42

19

86

19

08

19

08

19

86

21

42

29

10

22

08

25

20

29

10

25

20

22

08

19

74

18

18

17

40

17

40

18

18

19

74

30

78

23

76

26

88

30

78

26

88

23

76

21

42

19

86

19

08

19

08

19

86

21

42

83

05

2

: Picker Transport Time Table for Level B, C, D

Order Number

1

2

3

4

5

6

7

8

9

10

11

12

13

Order Number LA_Other LA_Footware

1 0

2 13

3 0

4 0

5 24

6 0

7 0

8 0

9 1

10 19

11 0

12 0

13 0

14 0

15 0

Table 17: Extraction from Total Transport Time for Each Level in Picking

Table 18: Extraction from Quantity

75

A B C D Grand Total

0 0 0 75.5 75.5

33.5 0 0 0 33.5

93.5 0 0 0 93.5

59.5 0 0 0 59.5

59.5 0 0 0 59.5

0 0 74.5 59.5 134

0 63.5 0 0 63.5

0 75.5 0 0 75.5

0 49.5 0 0 49.5

63.5 0 75.5 0 139

0 101.5 0 0 101.5

0 0 49.5 0 49.5

37.5 0 0 0 37.5

LA_Footware LB_Other LB_Footware LC_Other LC_Footware LD_Other

0 0 0 0 0

0 0 12 0 0

12 0 0 0 24

15 0 0 0 0

0 0 0 0 0

0 0 0 134 0

0 11 0 0 0

0 0 30 0 0

0 2 0 0 9

0 0 0 0 0

0 24 0 0 0

0 0 0 0 43

3 0 0 0 0

1 0 0 0 0

0 1 0 108 0

Total Transport Time for Each Level in Picking

Quantity and Business Units per Order on each Level

LD_Other LD_Footware

7 0

0 0

0 0

0 0

0 0

0 0

0 0

0 5

0 0

0 0

0 0

0 0

0 0

0 0

0 0

Order Number A

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Order Number

Table 19: Extraction from Number to be batched on a Level

Table 20: Extraction Distance the Conveyor travel on a Level

76

B C D Batch

0 0 0 1 1

1 1 0 0 2

1 0 1 0 2

1 0 0 0 1

1 0 0 0 1

0 0 1 0 1

0 1 0 1 2

0 1 0 0 1

1 1 1 0 3

1 0 0 0 1

0 1 0 0 1

0 0 1 0 1

1 0 0 0 1

1 0 0 0 1

0 1 1 0 2

Order Number A B C D

1 0 0 0 61

2 30 0 0 0

3 54 0 0 0

4 36 0 0 0

5 36 0 0 0

6 0 0 66 0

7 0 43 0 0

8 0 61 0 0

9 0 52 0 0

10 18 0 0 0

11 0 70 0 0

12 0 0 52 0

13 24 0 0 0

14 36 0 0 0

15 0 29 0 0

Number to be batched on a Level

Distance the Conveyor travel on a Level

10.2 Probability Distributions

10.2.1 Picking Process Time

Distribution Summary

Distribution: Triangular

Expression: TRIA (1, 19, 48)

Square Error: 0.003419

Data Summary

Number of Data Points = 13

Min Data Value = 1

Max Data Value = 48

Sample Mean = 19

Sample Std Dev = 8.8

Histogram Summary

Histogram Range = 1 to

Number of Intervals = 11

77

Distributions

Picking Process Time for Business Unit “Other” Distribution

= 130

= 1

48

19

= 8.8

1 to 48

11

10.2.2 Picking Process Time for Business Unit


Distribution: Triangular

Expression: TRIA (3, 11, 26)


Data Summary


Min Data Value = 3

Max Data Value = 26

Sample Mean = 11


Histogram Summary

Histogram Range = 3


78

Picking Process Time for Business Unit “Footwear” Distribution

= 130

= 3

= 26

= 11

= 4.1

= 3 to 26

= 11

Distribution

10.2.3 Packing Process Tim


Distribution: Normal

Expression: NORM (287, 98.2


Data Summary


Min Data Value = 0.1

Max Data Value = 542

Sample Mean = 287


Histogram Summary

Histogram Range = 0.1 to 542


79

Packing Process Time Distribution

, 98.2)

= 150

= 0.1

= 542

= 287

= 98.2

= 0.1 to 542

= 12

10.3 Simulation Results

10.3.1 Queue average waiting time per entity

Waiting Time in Queue (Avg)

Increase Growth

Base

Model

Scenario

Batch Order same

OrderNumber 6.122 6.376

Packing Station A 0

Packing Station B 0

Packing Station C 0

Packing Station D 0

Packing Station E 0

Packing Station F 0

Packing Station G 0

Packing Station H 0

Packing Station I 0

Packing Station J 0

Transport

between Stations

LA

76.407 82.43

Transport

between Stations

LB

54.37 58.392

Transport

between Stations

LC

39.336 42.666

Transport

between Stations

LD

35.659 38.735

80

Queue average waiting time per entity

Waiting Time in Queue (Avg)

Increase Growth Change #Resources

Scenario

2

Scenario

3

Scenario

1

Scenario

2

Scenario

3

6.376 6.767 4.535 3.754 5.654

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 3.207

0 0 0 0 N/A

0 0 0 0 N/A

0 0 1.047 6.161 N/A

82.43 87.56 55.787 39.232 75.108

58.392 61.944 39.894 27.526 53.391

42.666 46.654 30.353 20.549 39.652

38.735 41.892 27.563 18.953 36.017

Change

#Units

send

Scenario

Scenario 1

5.654 3.76

0

0

0

0

0

0

3.207 0

N/A 0

N/A 0

N/A 0

75.108 41.974

53.391 32.145

39.652 25.499

36.017 24.058

10.3.2 Number of entities entering

Number in Process (Avg)

Increase Growth

Base

Model

Scenario

Packing Station A 841 917

Packing Station B 704 737

Packing Station C 548 533

Packing Station D 362 317

Packing Station E 202 176

Packing Station F 102

Packing Station G 34

Packing Station H 14

Packing Station I 3

Packing Station J 0

Picking Process LA 1430 1430

Picking Process LB 669 669

Picking Process LC 393 393

Picking Process LD 318 318

81

Number of entities entering a process

Increase Growth Change #Resources

Scenario

2

Scenario

3

Scenario

1

Scenario

2

Scenario

3

917 937 663 504 831

737 773 592 481 692

533 538 511 431 547

317 312 388 378 352

176 151 267 297 215

87 69 180 246 105

31 21 104 176 68

12 8 57 120 N/A

0 1 32 80 N/A

0 0 16 97 N/A

1430 1430 1430 1430 1430

669 669 669 669 669

393 393 393 393 393

318 318 318 318 318

Change #Units

send

Scenario Scenario 1

856

690

526

340

216

114

49

17

2

0

1430 1430

669

393

318

10.4 The Simulation Parts

10.4.1 Order information generation

82

The Simulation Parts

Order information generation phase

10.4.2 Pick Face

83

10.4.3 Pack Station

84


85

Despatch and Output transfer phase

analysis of nike distribution facility’s outbound process flow

Documents